release: bump version to 0.41.1

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

.claude/commands/cleanup.md

@@ -0,0 +1,120 @@
+---
+description: 审查当前工作区未提交代码中的垃圾代码，并在不影响逻辑的前提下自动清理
+argument-hint: 可选：指定要检查的文件或目录（默认检查所有未提交修改）
+allowed-tools: ["Read", "Edit", "Bash", "Grep", "Glob"]
+---
+
+# /cleanup — 垃圾代码审查与清理
+
+分析当前工作区（git diff）中的未提交代码，找出并修复常见垃圾代码，**不得改变任何运行逻辑**。
+
+## 检查范围
+
+若 `$ARGUMENTS` 非空，则只检查指定文件/目录；否则检查所有未提交修改（`git diff HEAD`）。
+
+## 审查清单
+
+按优先级检查以下问题（只报告在本次 diff 中**新增或修改**的代码里存在的问题）：
+
+### 1. 重复逻辑 (Duplicate Logic)
+- 完全相同或高度相似的代码块在多处出现
+- 同一函数/方法被多个地方各自实现，已有公共版本未被复用
+- 相同的 DOM 查询、正则、模板字符串在同一文件重复
+
+### 2. Magic Numbers / Magic Strings
+- 裸数字直接参与计算（如偏移量、时间、尺寸、阈值），没有命名常量
+- 硬编码字符串（如 id 名、状态值、URL 片段）散落在逻辑中
+- 例外：`0`, `1`, `-1`, `100`, `""` 等语义明确的惯用值不算
+
+### 3. 命名问题
+- 含义不明的缩写变量（如 `or_`, `tmp2`, `x2`）
+- 命名与实际用途不符
+- 同一概念在不同地方用不同名字表达
+
+### 4. 死代码 / 无效代码
+- 注释掉的旧代码块（3行以上）
+- 声明后从未使用的变量/参数/导入
+- 永远不会执行的条件分支
+
+### 5. 代码风格问题
+- 尾部空白字符（trailing whitespace）
+- 同一文件内风格不一致（如混用单双引号、缩进不统一）
+- 空行使用不一致（连续多个空行等）
+
+### 6. 其他常见问题
+- 私有辅助函数应被 export 但没有，导致调用方重复实现
+- 类型/接口重复定义
+- 过于冗长的条件表达式可以简化（不改逻辑）
+
+## 执行步骤
+
+### Step 1 — 获取待检查文件列表
+
+```bash
+# 无参数时：获取所有未提交修改
+git diff HEAD --name-only
+
+# 有参数时：用 $ARGUMENTS 过滤
+```
+
+### Step 2 — 逐文件阅读并分析
+
+- 用 Read 工具读取完整文件（不只读 diff）
+- 对照审查清单，记录每个问题：文件名、行号、问题类型、建议修复方式
+
+### Step 3 — 报告问题清单
+
+在修改前，先以列表形式输出所有发现的问题：
+
+```
+发现 N 个问题：
+
+[文件] js/foo.js
+  · L34, L78: 重复逻辑 — 两处都实现了相同的 DOM 查询，可提取到 getPanel()
+  · L91: Magic number — 硬编码 14 作为偏移量，应命名为 TOOLTIP_OFFSET
+
+[文件] js/bar.js
+  · L12: 命名问题 — 变量 `or_` 语义不明，应命名为 outerR/outerG/outerB
+  ...
+```
+
+如果没有发现问题，直接输出"未发现垃圾代码，当前代码质量良好。"并停止。
+
+### Step 4 — 执行修复
+
+对每个问题，使用 Edit 工具进行**最小化修改**：
+
+- **重复逻辑**：提取为共享常量/函数，更新所有调用点
+- **Magic number**：在文件顶部或逻辑附近声明 `const NAME = value`，替换所有引用
+- **命名问题**：重命名变量，更新所有使用处
+- **死代码**：直接删除
+- **尾部空白/风格**：修正
+- **未 export 的函数**：添加 `export`，在调用方改为导入（不重复实现）
+
+**修复原则：**
+- 只改在审查清单中发现的问题，不做额外优化
+- 每次 Edit 只修改确实有问题的行，保持 diff 最小
+- 改完后用 `grep` 验证旧的坏代码已消失
+
+### Step 5 — 输出总结
+
+```
+清理完成：
+
+修复了 N 个问题：
+  ✓ earth.js — 提取重复 vertexShader 为 ATMOS_VERTEX_SHADER 常量
+  ✓ main.js — 提取 TOOLTIP_CURSOR_OFFSET = 14（4处引用）
+  ✓ controls.js — export updateLayerButtonState，移除 main.js 中的重复实现
+  ...
+
+未修改的问题（需人工确认）：
+  ! foo.js L45 — 注释代码块较长，建议手动确认是否可删除
+```
+
+## 约束
+
+- **禁止**改变函数签名、接口定义、导出 API（除非问题正是私有函数应被 export）
+- **禁止**添加新功能、新抽象、新参数
+- **禁止**修改注释内容（只删除注释掉的死代码）
+- **禁止**修改测试文件逻辑
+- 如果一个 Magic number 的语义不完全确定，**跳过**，在总结中标记为"需人工确认"

.claude/commands/goal-driven.md

@@ -0,0 +1,91 @@
+---
+description: 用 goal-driven 方法推动一个复杂任务持续执行，直到明确成功标准被满足
+argument-hint: 建议填写任务目标；若同时给出成功标准更好
+allowed-tools: ["Read", "Edit", "Bash", "Grep", "Glob"]
+---
+
+# /goal-driven — 目标驱动执行模式
+
+使用 `lidangzzz/goal-driven` 的核心思想来推进复杂任务：先固定目标与成功标准，再持续执行和反复验收，直到标准真正满足。
+
+适用场景：
+
+- 长周期实现任务
+- 高复杂度工程任务
+- 可被明确验收的研究、实现、迁移、验证类工作
+
+不适用场景：
+
+- 纯脑暴
+- 无法定义成功标准的模糊任务
+- 很小的一次性修改
+
+## 输入要求
+
+若 `$ARGUMENTS` 只包含目标，没有成功标准，先补全一版可执行的成功标准再开始。
+
+启动时先输出：
+
+```md
+Goal
+- ...
+
+Criteria for success
+- ...
+
+Plan
+1. ...
+2. ...
+3. ...
+
+Verification
+- ...
+```
+
+## 执行规则
+
+1. 先把任务固化为两个核心块：
+   - `Goal`
+   - `Criteria for success`
+
+2. 成功标准必须尽量客观，可验证，可落地。
+   优先写成：
+   - 需要交付什么
+   - 需要通过哪些测试或验证
+   - 如何判断结果真的完成
+
+3. 进入持续执行循环：
+   - 完成一个阶段
+   - 检查当前结果是否满足成功标准
+   - 若未满足，明确剩余差距并继续推进
+
+4. 任何“完成了”“差不多了”“已实现”之类的结论，都必须经过验证，不能直接接受。
+
+5. 如果验证失败：
+   - 明确指出哪条成功标准没满足
+   - 继续工作，不要把阶段性进展误判为完成
+
+6. 只有在以下情况之一才能停止：
+   - 成功标准已满足
+   - 用户明确要求停止
+
+## 执行风格
+
+- 重证据，轻口头判断
+- 重验收，轻自我感觉
+- 优先用测试、日志、产物、对比结果来证明完成
+- 对长期任务保持“未达标就继续”的节奏
+
+## 简版模板
+
+```md
+Goal: [[[[[在此填写最终目标]]]]]
+
+Criteria for success: [[[[[在此填写成功标准]]]]]
+
+循环执行：
+1. 推进任务
+2. 检查是否满足成功标准
+3. 若未满足，继续工作
+4. 直到满足标准或用户明确停止
+```

.claude/commands/release.md

@@ -0,0 +1,146 @@
+---
+description: 发版工作流：根据变更类型决定版本号，更新所有版本文件和 changelog，运行验证，commit 并 push
+argument-hint: 可选：feature | bugfix | 或直接描述本次发布内容
+allowed-tools: ["Read", "Edit", "Bash", "Glob", "Grep"]
+---
+
+# /release — Planet 发版工作流
+
+## 版本号规则
+
+| 变更类型 | 版本跳动 | 适用场景 |
+|---------|---------|---------|
+| `feature` | `+0.1.0` | 纯新功能，无 bugfix |
+| `improvement` | `+0.0.1` | UI 调整、小功能增强、bugfix 混合，或以 UI/体验改进为主的迭代 |
+| `bugfix` | `+0.0.1` | 纯 bug 修复，无新功能 |
+| `docs` / `maintenance` / `refactor` | 默认不发版，除非用户明确要求 |
+
+意图混合时以用户明确描述为准；bugfix + 小 feature 混合默认判定为 `improvement`（`+0.0.1`）。
+
+## 必须同步更新的文件
+
+使用 `git rev-parse --show-toplevel` 获取仓库根目录，以下路径均相对于根目录：
+
+- `VERSION`
+- `frontend/package.json`（`"version"` 字段）
+- `pyproject.toml`（`version =` 字段）
+- `uv.lock`（**不要手动编辑**，通过 `uv lock` 重新生成）
+- `docs/CHANGELOG.md`
+- `docs/version-history.md`
+
+## 执行步骤
+
+### Step 1 — 环境检查
+
+```bash
+git branch --show-current          # 确认在 dev 分支
+git status --short                 # 检查是否有无关的未暂存修改
+cat VERSION                        # 读取当前版本
+```
+
+若当前**不在 `dev` 分支**，停下来告知用户，不要继续。
+
+若存在无关的未暂存修改，列出并询问用户是否一并提交，或先 stash。
+
+### Step 2 — 确定发版类型与新版本号
+
+- 若 `$ARGUMENTS` 提供了明确类型（`feature` / `bugfix`），直接使用
+- 否则根据当前 `git diff HEAD` 和 `git log` 推断
+- 计算新版本号（例：`0.26.2` → bugfix → `0.26.3`）
+- **先输出发版计划供用户确认**：
+
+```
+发版计划：
+  类型：bugfix
+  版本：0.26.2 → 0.26.3
+  分支：dev
+  将更新：VERSION, frontend/package.json, pyproject.toml, uv.lock, CHANGELOG.md, version-history.md
+```
+
+### Step 3 — 更新版本号文件
+
+按顺序更新（每步用 Edit 工具，精确替换，不要重写整个文件）：
+
+1. `VERSION` — 直接替换全部内容为新版本号
+2. `frontend/package.json` — 替换 `"version": "x.x.x"` 行
+3. `pyproject.toml` — 替换 `version = "x.x.x"` 行
+4. 运行 `uv lock` 重新生成 `uv.lock`（在仓库根目录下执行）
+
+### Step 4 — 更新 CHANGELOG.md
+
+在文件顶部插入新条目，格式：
+
+```markdown
+## [x.x.x] — YYYY-MM-DD
+
+### ✨ Features / 🐛 Fixes / 🔧 Improvements
+- ...（只列高信号条目，最多 5 条）
+- ...
+
+---
+```
+
+日期使用 `date +%Y-%m-%d` 获取今天的日期。
+
+### Step 5 — 更新 docs/version-history.md
+
+- 更新文件头部的"当前开发版本"字段
+- 在时间线表格顶部插入新行：`| vx.x.x | YYYY-MM-DD | 一句话摘要 |`
+
+### Step 6 — 验证
+
+针对本次变更范围做最小验证：
+
+- Python 文件有修改：`python3 -m py_compile <changed_files>`
+- Frontend 文件有修改：运行项目标准检查（若无则跳过并说明）
+- 版本号一致性检查：用 grep 确认 VERSION、package.json、pyproject.toml 中的版本号完全一致
+
+```bash
+grep -h "version" VERSION frontend/package.json pyproject.toml
+```
+
+### Step 7 — 提交前预览
+
+展示将要提交的文件列表：
+
+```bash
+git diff --stat HEAD
+```
+
+再次确认所有必须文件都在变更列表中，**不包含**非预期文件（如调试文件、.env 等）。
+
+### Step 8 — Commit & Push（用户确认后）
+
+```bash
+git add VERSION frontend/package.json pyproject.toml uv.lock docs/CHANGELOG.md docs/version-history.md
+# 若有代码变更也一并 stage
+git add <code_files>
+
+git commit -m "release: bump version to x.x.x"
+git tag vx.x.x
+git push origin dev
+git push origin vx.x.x
+```
+
+commit message 固定格式：`release: bump version to x.x.x`
+
+### Step 9 — 完成确认
+
+输出摘要：
+
+```
+✓ 版本号已更新：0.26.2 → 0.26.3
+✓ CHANGELOG 已更新
+✓ version-history 已更新
+✓ uv.lock 已重新生成
+✓ 验证通过
+✓ commit: release: bump version to 0.26.3
+✓ tag: v0.26.3
+✓ 已 push 到 origin/dev
+```
+
+## 注意事项
+
+- `uv.lock` 只能通过 `uv lock` 生成，绝不手动编辑
+- 发版 commit 只包含版本文件 + 本次功能代码，不混入无关改动
+- 若环境中 `uv` 不可用，说明原因并跳过 lockfile 更新，提醒用户手动运行

.codex/config.toml

@@ -0,0 +1,3 @@
+approval_policy = "never"
+
+sandbox_mode = "danger-full-access"

.codex/skills/cleanup/SKILL.md

@@ -0,0 +1,124 @@
+---
+name: cleanup
+description: Use when the user asks to clean up, lint, or review uncommitted code for common code smells — duplicate logic, magic numbers, unclear naming, dead code, style inconsistencies. Fixes issues without changing any runtime behavior.
+---
+
+# Cleanup
+
+Review and fix code quality issues in the current working tree without altering any logic or behavior.
+
+## When To Use
+
+- The user asks to clean up, tidy, or lint uncommitted changes
+- The user wants a code smell review before releasing or committing
+- The user mentions magic numbers, duplicate logic, dead code, or naming issues
+
+Do not refactor architecture, add features, or change behavior.
+
+## Scope
+
+If the user specifies a file or directory, check only that. Otherwise check all uncommitted changes (`git diff HEAD`).
+
+Only report issues present in **newly added or modified** lines of this diff — do not audit unchanged code.
+
+## Checklist
+
+### 1. Duplicate Logic
+- Identical or near-identical code blocks appearing in multiple places
+- A function/helper that already exists but is re-implemented elsewhere instead of being reused
+- Repeated DOM queries, regex literals, or template strings within the same file
+
+### 2. Magic Numbers / Magic Strings
+- Bare numeric literals used in calculations (offsets, timeouts, sizes, thresholds) without a named constant
+- Hardcoded strings (IDs, status values, URL fragments) scattered through logic
+- Exceptions: `0`, `1`, `-1`, `100`, `""` and other idiomatically clear values are fine
+
+### 3. Naming Issues
+- Cryptic abbreviations (`or_`, `tmp2`, `x2`)
+- Names that do not match actual behavior
+- The same concept referred to by different names in different places
+
+### 4. Dead Code
+- Commented-out code blocks (3+ lines)
+- Variables, parameters, or imports declared but never used
+- Branches that can never execute
+
+### 5. Style Inconsistencies
+- Trailing whitespace
+- Mixed quote styles or indentation within the same file
+- Inconsistent blank-line usage (multiple consecutive blank lines, etc.)
+
+### 6. Other
+- Private helper functions that should be exported but are not, causing callers to duplicate the implementation
+- Overly verbose conditions that can be simplified without changing logic
+
+## Steps
+
+### Step 1 — Get the file list
+
+```bash
+git diff HEAD --name-only
+```
+
+Filter to the user-specified path if one was provided.
+
+### Step 2 — Read and analyze each file
+
+Read the full file (not just the diff) with the Read tool. For each file, record every issue found: filename, line number, category, and suggested fix.
+
+### Step 3 — Report findings before touching anything
+
+Print a structured list:
+
+```
+Found N issues:
+
+[file] js/foo.js
+  · L34, L78: Duplicate logic — same DOM query implemented twice; extract to getPanel()
+  · L91: Magic number — bare 14 used as pixel offset; name it TOOLTIP_OFFSET
+
+[file] js/bar.js
+  · L12: Naming — variable `or_` is unclear; rename to outerR, outerG, outerB
+  ...
+```
+
+If no issues are found, output "No code smells detected. Code quality looks good." and stop.
+
+### Step 4 — Fix each issue
+
+Use the Edit tool for **minimal, targeted changes**:
+
+- **Duplicate logic**: extract to a shared constant or function; update all call sites
+- **Magic number/string**: declare `const NAME = value` near the top of the relevant scope; replace all usages
+- **Naming**: rename the variable/function; update all references
+- **Dead code**: delete it
+- **Trailing whitespace / style**: fix in place
+- **Unexported helper**: add `export`; update callers to import instead of re-implementing
+
+Principles:
+- Only fix issues identified in the checklist — no extra improvements
+- Keep each Edit as small as possible
+- After fixing, verify the old bad pattern is gone with grep
+
+### Step 5 — Summary
+
+```
+Cleanup complete:
+
+Fixed N issues:
+  ✓ earth.js — extracted duplicate vertexShader into ATMOS_VERTEX_SHADER constant
+  ✓ main.js — extracted TOOLTIP_CURSOR_OFFSET = 14 (4 references updated)
+  ✓ controls.js — exported updateLayerButtonState; removed duplicate implementation in main.js
+  ...
+
+Skipped (needs manual review):
+  ! foo.js L45 — large commented-out block; confirm it is safe to delete
+```
+
+## Constraints
+
+- **Do not** change function signatures, exported interfaces, or public APIs (unless the issue is a missing export)
+- **Do not** add new features, abstractions, or parameters
+- **Do not** rewrite comments (only delete commented-out dead code)
+- **Do not** touch test file logic
+- If a magic number's intent is uncertain, skip it and flag it in the summary

.codex/skills/goal-driven/SKILL.md

@@ -0,0 +1,101 @@
+---
+name: goal-driven
+description: Run a goal-driven execution loop for very large, long-horizon, rigorously verifiable tasks. Use when the user explicitly wants the lidangzzz/goal-driven method, a master-agent plus worker-agent style workflow, or a persistent loop that keeps working until concrete success criteria are satisfied.
+---
+
+# Goal-Driven
+
+Use this skill when the user wants a strict goal-driven workflow for a hard task with:
+
+- one clear end goal
+- explicit success criteria
+- repeated verification against those criteria
+- continued execution until the criteria are actually met
+
+This skill is adapted from `lidangzzz/goal-driven`, but trimmed for local skill use to avoid bloating context.
+
+## When To Use
+
+Use it for tasks like:
+
+- compilers, interpreters, theorem-like proof work, deep refactors
+- long-running system design or implementation work
+- problems that are expensive and complex, but still objectively testable
+
+Do not use it for:
+
+- vague brainstorming without a success condition
+- short one-shot edits
+- tasks where "done" cannot be evaluated in a meaningful way
+
+## Core Model
+
+The workflow has two roles:
+
+1. Master role
+   Defines the goal, defines the success criteria, audits progress, and decides whether the work is actually complete.
+
+2. Worker role
+   Keeps advancing the task toward the goal. If a result is partial, stalled, or unverifiable, the worker continues.
+
+In Codex, only use actual subagents when the user explicitly asks for delegation or subagent work and the platform supports it. Otherwise emulate the same loop locally: keep working, checkpointing, and re-verifying until the criteria are satisfied.
+
+## Workflow
+
+1. Normalize the task into two blocks:
+   - `Goal`
+   - `Criteria for success`
+
+2. Make the criteria concrete and testable.
+   Good criteria usually include:
+   - required outputs
+   - required validations or tests
+   - edge cases or coverage thresholds
+   - what evidence proves completion
+
+3. Break the work into milestones that can each produce evidence.
+
+4. Execute the next milestone.
+   If subagents are explicitly allowed, the master may delegate bounded worker tasks.
+   If not, do the work locally but keep the master/worker mindset.
+
+5. Whenever work pauses, stalls, or appears complete, audit against the criteria directly.
+   Check artifacts, tests, logs, diffs, metrics, or other real evidence.
+
+6. If the criteria are not met, continue with a specific delta:
+   - what is still missing
+   - what evidence failed
+   - what the next worker pass must improve
+
+7. Stop only when the criteria are met, or when the user explicitly stops the process.
+
+## Operating Rules
+
+- Prefer objective checks over self-reported completion.
+- Do not confuse progress with completion.
+- If the worker says "done", verify it.
+- If verification fails, continue from the gap instead of restarting blindly.
+- Keep the goal stable unless the user changes it.
+- Tighten fuzzy criteria before sinking large amounts of effort.
+
+## Recommended Response Shape
+
+When starting a goal-driven task, structure the kickoff like this:
+
+```md
+Goal
+- ...
+
+Criteria for success
+- ...
+
+Current plan
+1. ...
+2. ...
+3. ...
+
+Verification
+- What evidence will prove completion
+```
+
+For a reusable prompt template, read [references/prompt-template.md](references/prompt-template.md).

.codex/skills/goal-driven/agents/openai.yaml

@@ -0,0 +1,7 @@
+interface:
+  display_name: "Goal-Driven"
+  short_description: "Drive complex work until explicit success criteria are met."
+  default_prompt: "Use $goal-driven to turn this task into a concrete goal, explicit success criteria, and a verification-driven execution loop."
+
+policy:
+  allow_implicit_invocation: true

.codex/skills/goal-driven/references/prompt-template.md

@@ -0,0 +1,38 @@
+# Goal-Driven Prompt Template
+
+Use this when you want a reusable kickoff prompt for a master/worker execution loop.
+
+```md
+# Goal-Driven System
+
+Goal: [[[[[DEFINE THE FINAL GOAL HERE]]]]]
+
+Criteria for success: [[[[[DEFINE THE SUCCESS CRITERIA HERE]]]]]
+
+You are the master agent.
+
+Your job is to:
+1. Keep the goal and criteria fixed.
+2. Start worker execution toward the goal.
+3. Audit any claimed progress against the criteria.
+4. If the criteria are not met, continue the work with a precise next delta.
+5. Stop only when the criteria are satisfied or the user explicitly stops the process.
+
+Worker requirements:
+1. Break the task into subproblems.
+2. Keep producing concrete progress toward the goal.
+3. Report evidence, not just claims.
+4. Continue until the criteria are satisfied.
+
+Master audit loop:
+1. Check whether the worker is still making progress.
+2. If the worker stalls or claims completion, verify against the criteria.
+3. If verification fails, resume work from the remaining gap.
+4. Repeat until the criteria are met.
+```
+
+## Notes
+
+- Stronger criteria produce better results than stronger rhetoric.
+- Prefer measurable checks such as tests, parity checks, generated artifacts, benchmarks, or reviewable outputs.
+- If the environment does not support subagents, emulate the same loop locally.

.codex/skills/release/SKILL.md

@@ -0,0 +1,157 @@
+---
+name: release
+description: Use when the user asks to release, bump version, update changelog/version files, or commit/push a repository release for the Planet repo. Determines version bump type from changes, updates all required version-bearing files, updates changelog and version-history, runs minimal validation, then commits, tags, and pushes.
+---
+
+# Release Workflow
+
+Use this skill for release-oriented work in this repository.
+
+## When To Use
+
+- The user asks to `发版`
+- The user asks to bump a version
+- The user asks to update `CHANGELOG`, `version-history`, or version files as part of a release
+- The user asks to commit/push a release or a publishable bugfix/feature bundle
+
+Do not use this skill for ordinary commits that are not being released.
+
+## Versioning Rules
+
+- `feature` -> bump `+0.1.0`
+- `bugfix` -> bump `+0.0.1`
+- `docs`, `maintenance`, and `refactor` do not bump by default unless the user explicitly wants a release
+
+When intent is mixed, prefer the user's stated release intent.
+
+## Required Files
+
+Use `git rev-parse --show-toplevel` to get the repo root. All paths are relative to it:
+
+- `VERSION`
+- `frontend/package.json` (`"version"` field)
+- `pyproject.toml` (`version =` field)
+- `uv.lock` (**never edit manually** — regenerate by running `uv lock`)
+- `docs/CHANGELOG.md`
+- `docs/version-history.md`
+
+## Workflow
+
+### Step 1 — Environment check
+
+```bash
+git branch --show-current   # must be on dev
+git status --short           # check for unrelated uncommitted changes
+cat VERSION                  # read current version
+```
+
+If not on `dev`, stop and tell the user. Do not proceed.
+
+If unrelated uncommitted changes exist, list them and ask the user whether to include them or stash first.
+
+### Step 2 — Determine release type and next version
+
+- If the user provided an explicit type (`feature` / `bugfix`), use it
+- Otherwise infer from `git diff HEAD` and recent `git log`
+- Compute the next version (e.g. `0.26.2` → bugfix → `0.26.3`)
+- **Show the release plan before making any changes:**
+
+```
+Release plan:
+  Type:    bugfix
+  Version: 0.26.2 → 0.26.3
+  Branch:  dev
+  Will update: VERSION, frontend/package.json, pyproject.toml, uv.lock, CHANGELOG.md, version-history.md
+```
+
+### Step 3 — Update version files
+
+Update in order (use Edit for precise replacement, never rewrite whole files):
+
+1. `VERSION` — replace entire content with new version string
+2. `frontend/package.json` — replace `"version": "x.x.x"` line
+3. `pyproject.toml` — replace `version = "x.x.x"` line
+4. Run `uv lock` at repo root to regenerate `uv.lock`
+
+### Step 4 — Update CHANGELOG.md
+
+Insert a new entry at the top of the file:
+
+```markdown
+## x.x.x
+
+Released: YYYY-MM-DD
+
+### Highlights
+
+- ...
+
+### Added / Fixed / Improved
+
+- ... (high-signal items only, max 5)
+
+---
+```
+
+Get today's date with `date +%Y-%m-%d`.
+
+### Step 5 — Update docs/version-history.md
+
+- Update the "current dev version" field in the file header
+- Insert a new row at the top of the timeline table: `| vx.x.x | YYYY-MM-DD | one-line summary |`
+
+### Step 6 — Validate
+
+Run the smallest relevant validation for the changes in scope:
+
+- Python files changed: `python3 -m py_compile <changed_files>`
+- Frontend files changed: run the project-standard check if available; otherwise skip and say so
+- Version consistency: confirm VERSION, package.json, pyproject.toml, and uv.lock all show the same version
+
+```bash
+grep -h "version" VERSION frontend/package.json pyproject.toml
+```
+
+### Step 7 — Pre-commit preview
+
+Show what will be committed:
+
+```bash
+git diff --stat HEAD
+```
+
+Confirm all required files are present and no unexpected files (debug files, `.env`, etc.) are included.
+
+### Step 8 — Commit, tag, and push
+
+```bash
+git add VERSION frontend/package.json pyproject.toml uv.lock docs/CHANGELOG.md docs/version-history.md
+# also stage any code changes included in this release
+git add <code_files>
+
+git commit -m "release: bump version to x.x.x"
+git tag vx.x.x
+git push origin dev
+git push origin vx.x.x
+```
+
+Commit message format is fixed: `release: bump version to x.x.x`
+
+### Step 9 — Completion summary
+
+```
+✓ Version bumped: 0.26.2 → 0.26.3
+✓ CHANGELOG updated
+✓ version-history updated
+✓ uv.lock regenerated
+✓ Validation passed
+✓ commit: release: bump version to 0.26.3
+✓ tag: v0.26.3
+✓ Pushed to origin/dev
+```
+
+## Notes
+
+- `uv.lock` must only be updated by running `uv lock`, never manually
+- The release commit should include only version files + the code for this release — no unrelated changes
+- If `uv` is unavailable in the environment, say so explicitly and remind the user to run it manually

.gitignore

@@ -145,3 +145,8 @@ docs/.venv/
 *.temp
 tmp/
 temp/
+
+# ----------------------
+# Runtime Data
+# ----------------------
+data/ai/bgp-briefs/

README.md

@@ -102,6 +102,13 @@
 | Axios | HTTP 客户端 |
 | Socket.io-client | WebSocket 客户端 |
 | ECharts | 统计图表 |
+| Bun | 前端包管理与脚本运行 |
+
+前端工程统一使用 Bun：
+
+- 安装依赖使用 `bun install`
+- 运行脚本使用 `bun run <script>`
+- 不使用 `npm`、`pnpm`、`yarn`
 
 ### 虚幻引擎客户端
 
@@ -205,10 +212,161 @@
 ./planet.sh health
 ```
 
+前端命令约定：
+
+```bash
+cd frontend
+bun install
+bun run dev
+bun run build
+```
+
+不要使用 `npm run ...`，避免在 WSL/Windows 混合环境里触发 `cmd.exe` 路径兼容问题。
+
 ## API 文档
 
 启动服务后访问: `http://localhost:8000/docs`
 
+## WSL / Windows 局域网访问
+
+如果服务运行在 WSL 中，而你希望：
+
+- Windows 本机浏览器访问开发服务
+- 同一局域网内的手机或其他电脑访问开发服务
+
+推荐按下面顺序排查和配置。
+
+### 1. 在 WSL 中启动服务
+
+```bash
+./planet.sh start --allow-lan
+```
+
+这会让前端监听 `0.0.0.0:3000`，后端监听 `0.0.0.0:8000`。
+
+### 2. 先确认 WSL 内部服务正常
+
+在 WSL 中执行：
+
+```bash
+curl http://localhost:3000
+curl http://localhost:8000/health
+ss -ltnp | grep -E ':3000|:8000'
+```
+
+预期：
+
+- `3000` 返回前端 HTML
+- `8000/health` 返回健康检查 JSON
+- `ss` 中能看到 `0.0.0.0:3000` 和 `0.0.0.0:8000`
+
+如果这一步不通，先不要继续做 Windows 转发。
+
+### 3. 在 Windows 本机验证 localhost 直通
+
+在 Windows PowerShell 中执行：
+
+```powershell
+curl http://localhost:3000
+curl http://localhost:8000/health
+```
+
+在常见的 WSL2 开发环境下，Windows 通常可以直接通过 `localhost` 访问 WSL 中的服务。
+
+### 4. 如果需要让局域网设备访问，再做 Windows 端口转发
+
+注意：下面的命令必须在“以管理员身份运行”的 PowerShell 中执行。
+
+先把 Windows 对外网卡上的 `3000` / `8000` 转发到 Windows 本机 `127.0.0.1`：
+
+```powershell
+netsh interface portproxy delete v4tov4 listenaddress=0.0.0.0 listenport=3000
+netsh interface portproxy delete v4tov4 listenaddress=0.0.0.0 listenport=8000
+
+netsh interface portproxy add v4tov4 listenaddress=0.0.0.0 listenport=3000 connectaddress=127.0.0.1 connectport=3000
+netsh interface portproxy add v4tov4 listenaddress=0.0.0.0 listenport=8000 connectaddress=127.0.0.1 connectport=8000
+```
+
+再放行 Windows 防火墙：
+
+```powershell
+New-NetFirewallRule -DisplayName "WSL Planet 3000" -Direction Inbound -Action Allow -Protocol TCP -LocalPort 3000
+New-NetFirewallRule -DisplayName "WSL Planet 8000" -Direction Inbound -Action Allow -Protocol TCP -LocalPort 8000
+```
+
+检查转发规则是否生效：
+
+```powershell
+netsh interface portproxy show all
+```
+
+预期能看到：
+
+- `0.0.0.0:3000 -> 127.0.0.1:3000`
+- `0.0.0.0:8000 -> 127.0.0.1:8000`
+
+### 5. 查 Windows 局域网 IP，并让其他设备访问
+
+在 Windows PowerShell 中执行：
+
+```powershell
+ipconfig
+```
+
+找到当前联网网卡的 IPv4 地址，例如 `192.168.8.228`。
+
+局域网其他设备可访问：
+
+- `http://<Windows局域网IP>:3000/earth`
+- `http://<Windows局域网IP>:3000/admin`
+
+例如：
+
+- `http://192.168.8.228:3000/earth`
+
+### 6. 常见现象与判断
+
+- WSL 中 `curl localhost:3000` 能通，但 Windows 访问 `WSL 的局域网 IP:3000` 不通：这是正常现象之一，优先验证 Windows 的 `localhost:3000`
+- Windows `localhost:3000` 能通，但局域网设备访问 `Windows 局域网 IP:3000` 不通：通常缺少 `portproxy` 或防火墙放行
+- `whoami /groups` 中 `S-1-5-32-544` 显示 `deny only`：说明当前 PowerShell 不是提权管理员窗口
+
+### 7. 本项目一次性验证顺序
+
+建议固定按这个顺序验证：
+
+1. WSL 中执行 `curl http://localhost:3000`
+2. WSL 中执行 `curl http://localhost:8000/health`
+3. Windows 中执行 `curl http://localhost:3000`
+4. Windows 中执行 `curl http://localhost:8000/health`
+5. 管理员 PowerShell 配置 `portproxy` 和防火墙
+6. 用手机或其他电脑访问 `http://<Windows局域网IP>:3000/earth`
+
+## 启动容错参数
+
+`planet.sh` 现在为依赖安装、数据库、AI Provider 启动加入了有限次重试，并会在数据库与 `aiprovider` 启动后额外等待 Docker healthcheck。
+
+可通过环境变量临时调整:
+
+```bash
+# 例: 放宽 AI Provider 与数据库在网络抖动下的自愈次数
+AI_PROVIDER_START_MAX_RETRIES=5 \
+AI_PROVIDER_RETRY_INTERVAL=10 \
+DATABASE_START_MAX_RETRIES=5 \
+DATABASE_RETRY_INTERVAL=10 \
+./planet.sh restart
+```
+
+常用参数:
+
+- `DEPENDENCY_INSTALL_MAX_RETRIES` / `DEPENDENCY_INSTALL_RETRY_INTERVAL`: 控制 `uv sync`、`bun install` 的重试次数与间隔，默认 `3` 次、`5` 秒
+- `DATABASE_START_MAX_RETRIES` / `DATABASE_RETRY_INTERVAL`: 控制 `postgres`、`redis` 的启动/重启与健康检查自愈，默认 `3` 次、`5` 秒
+- `AI_PROVIDER_START_MAX_RETRIES` / `AI_PROVIDER_RETRY_INTERVAL`: 控制 `aiprovider` 的构建/启动与容器重启自愈，默认 `3` 次、`5` 秒
+- `BACKEND_MAX_RETRIES`: 控制后端进程启动重试次数，默认 `3`
+- `FRONTEND_MAX_RETRIES`: 控制前端 dev server 启动重试次数，默认 `3`
+- `BACKEND_HEALTH_CHECK_ATTEMPTS` / `BACKEND_HEALTH_CHECK_INTERVAL`: 控制后端 HTTP 健康检查等待次数与间隔，默认 `10` 次、`2` 秒
+- `FRONTEND_HEALTH_CHECK_ATTEMPTS` / `FRONTEND_HEALTH_CHECK_INTERVAL`: 控制前端 HTTP 可访问检查等待次数与间隔，默认 `10` 次、`2` 秒
+- `AI_PROVIDER_HEALTH_CHECK_ATTEMPTS` / `AI_PROVIDER_HEALTH_CHECK_INTERVAL`: 控制 `aiprovider` HTTP 健康检查等待次数与间隔，默认 `10` 次、`2` 秒
+
 ## AI 接口预留
 
 项目现在采用“两层”设计:
@@ -284,8 +442,25 @@ AI_PROVIDER_SERVICE_TOKEN=change_me
 
 详细文档:
 
-- [docs/aiprovider.md](/home/ray/dev/linkong/planet/docs/aiprovider.md)
+- [docs/technical/agents-aiprovider.md](/home/ray/dev/linkong/planet/docs/technical/agents-aiprovider.md)
 - [aiprovider/README.md](/home/ray/dev/linkong/planet/aiprovider/README.md)
+- [docs/technical/frontend-layout-guidelines.md](/home/ray/dev/linkong/planet/docs/technical/frontend-layout-guidelines.md)
+- [docs/plans/frontend-ai-playground-development-plan.md](/home/ray/dev/linkong/planet/docs/plans/frontend-ai-playground-development-plan.md)
+- [docs/plans/agents-situational-awareness-foundation-plan.md](/home/ray/dev/linkong/planet/docs/plans/agents-situational-awareness-foundation-plan.md)
+
+## 前端页面布局规范
+
+管理后台页面默认遵循“单屏工作区”原则：
+
+- 页头、摘要区、主工作区应在一屏内形成稳定结构
+- 主表格 / 主图表 / 主分析区应占据页面主要可视空间
+- 模块内容超出时优先在卡片、表格、标签页内部滚动
+- 不依赖整页纵向撑开来容纳主要工作区
+
+当前推荐参考实现：
+
+- [frontend/src/pages/BGP/BGP.tsx](/home/ray/dev/linkong/planet/frontend/src/pages/BGP/BGP.tsx)
+- [docs/technical/frontend-layout-guidelines.md](/home/ray/dev/linkong/planet/docs/technical/frontend-layout-guidelines.md)
 
 ## License
 

TODO.md

@@ -13,7 +13,25 @@
 - [x] 接入 `IPtoASN / IPtoCountry` 作为 prefix-centric geography 的主数据源
 - [x] 接入 `OpenGeoFeed` 作为 prefix geography 的高质量覆盖/override 数据源
 - [x] 把 RIR delegated 设计成 prefix geography 的 fallback，而不是主来源
+- [ ] 为 `aiprovider` 建立 `provider -> api adapter -> compat policy` 的配置中心，优先落成 `json` 或 `yaml` 文件，运行时按 `provider/model` 读取兼容设置，而不是把专项兼容继续散落在 Python 分支里
+- [ ] 为市面上主流 AI 服务补专项兼容配置并固化到配置文件中，至少覆盖 `OpenAI / Anthropic / MiniMax / Ollama / Moonshot / DeepSeek / Qwen / GLM / Gemini / OpenRouter / vLLM / LM Studio / One API`
+- [ ] 在兼容配置中补齐可声明项：`api adapter`、`base_url pattern`、`auth header`、`thinking default`、`reasoning block mapping`、`stream path`、`tool-call capability`、`multimodal capability`、`provider-specific request patch`
 - [ ] 接入 `inetnum` / `inet6num` whois 作为比 RIR 更细粒度的后备层
 - [x] 在 activity layer 之后继续补 `route leak` 和 `path instability / flap` detector
 - [ ] 对 [frontend/public/earth/js/bgp.js](/home/ray/dev/linkong/planet/frontend/public/earth/js/bgp.js) 做按职责拆分的小重构，拆成 data / markers / overlays / animation，降低后续维护复杂度
 - [ ] 可选优化（非必做）：将 BGP incident/collector 标点改为 HTML marker（参考 worldmonitor 的 `htmlElementsData` 思路），实现近乎固定屏幕尺寸与更高密度可点击性
+- [ ] 保持 Earth 当前这批纯个人偏好设置继续走本地持久化：`旋转模式`、HUD 面板显示/隐藏、`地形透明度` 暂不升级到后端系统设置，避免把设备级偏好过早做成全局配置
+- [ ] 如果后续明确需要“账号级同步 Earth 偏好”，再单独设计 `Earth user preferences`：优先按用户维度而不是全局系统设置保存，并规划 `localStorage -> backend` 的平滑迁移策略
+- [ ] 为 Planet / Earth 补一个可用的日志查看系统：先明确前后端/AI Provider/采集任务的日志入口、最近日志聚合、筛选与 tail 能力，再决定是先做脚本级统一入口还是控制台内置日志面板
+- [ ] 把 Earth 态势新闻源从 [earth_news.py](/home/ray/dev/linkong/planet/backend/app/services/earth_news.py) 的硬编码列表抽成可配置目录，优先保持当前“实时聚合”链路不变，只先解决新闻源不可配置的问题
+- [ ] 为 Earth 态势新闻设计后续采集器化方案：明确新闻数据模型、去重策略、区域映射、过期清理和 Earth/AI 复用方式，再决定何时把新闻从实时抓取升级成正式 collector
+- [ ] 为 Earth 地球表面增加一层与基础纹理对齐的材质/纹理 overlay，并在同层叠加国界轮廓参考线；要求国界线与底图稳定对齐，且 hover 到国家轮廓时能高亮当前国家，便于校准地表和增强交互
+- [ ] 把 Earth 新闻接入通用巡航队列：按新闻发生地和时间排序生成巡航目标，巡航聚焦到新闻事件时显示对应新闻卡片，并保持实现边界为“通用巡航层 + 新闻业务适配层”，不要再把新闻逻辑直接耦合回 `main.js` 状态机
+- [ ] 为未知位置的算力中心建立分层坐标补全链路：优先 `精确坐标 > 站点/园区命中 > 城市 > 州/省 > 国家内主要算力城市 > 国家质心`，并把每次回退的 `confidence / reason / precision` 明确写进统一 GeoJSON
+- [ ] 为算力中心补一份可维护的本地位置注册表，例如 `canonical_name / aliases / operator / country / region / city / lat / lon / confidence / source_note`，避免把地点知识长期硬编码在 `visualization.py`
+- [ ] 增强 `epoch_ai_gpu` 和相关算力采集器的源页面解析：即使公开 API 不给坐标，也继续尝试从详情页、HTML、内嵌 JSON、schema.org、OpenGraph、脚本变量和 PDF/新闻稿链接里抽地点线索
+- [ ] 为未知位置算力中心增加外部富化策略评估：可选接入公开知识源或搜索兜底，只抓“站点名/园区名/城市名”级别线索，不直接抓经纬度结论，并把结果作为候选证据而不是真值
+- [ ] 为算力中心建立 `operator / cluster name / facility alias` 归一化层，先解决 `xAI / Colossus / Memphis`、`OpenAI / Stargate`、`CoreWeave`、`Lambda`、`Crusoe` 这类同一对象多种写法导致的地点匹配失败
+- [ ] 为估算位置增加更细的视觉和产品表达：除了问号角标，还要支持 tooltip/详情中的“估算依据”“精度级别”“最后核验时间”，并允许在设置中单独开关“仅看精确位置”
+- [ ] 为国家级估算点设计更合理的落点策略：优先落在“该国主要算力/数据中心城市候选集”而不是几何质心，必要时同国多节点做稳定散列分配，避免大量节点堆在荒漠或海上
+- [ ] 为未知位置算力中心建立人工校验工作流：支持导出待核验清单、记录人工确认结果，并把人工确认反哺到位置注册表，逐步减少问号点比例

VERSION

@@ -1 +1 @@
-0.23.0
+0.41.1

aiprovider/.env.example

@@ -6,29 +6,51 @@ AI_TIMEOUT_SECONDS=60
 AI_HTTP_RETRY_ATTEMPTS=2
 AI_ANALYSIS_SYSTEM_PROMPT=你是态势感知分析助手。请基于输入的上下文、观测与约束，输出结构化、克制、可执行的分析。
 
-# Select one provider mode:
-# - openai_compatible
-# - claude_compatible
+# Provider identity. Recommended values:
+# - minimax
+# - openai
 # - ollama
-AI_PROVIDER=ollama
+# Compatibility aliases still accepted:
+# - openai_compatible
+# - anthropic_compatible
+# - claude_compatible
+AI_PROVIDER=minimax
+
+# Request adapter style, following OpenClaw's API-seam pattern:
+# - auto
+# - openai-completions
+# - anthropic-messages
+# - ollama-generate
+AI_PROVIDER_API=anthropic-messages
 
 # Common model selection
-AI_MODEL=qwen2.5:7b
+AI_MODEL=MiniMax-M2.7
+
+# MiniMax CN Anthropic-compatible example
+AI_BASE_URL=https://api.minimaxi.com/anthropic
+AI_API_KEY=sk-cp-change-me
+AI_MAX_TOKENS=1200
+AI_ANTHROPIC_VERSION=2023-06-01
 
 # OpenAI-compatible example (vLLM / LM Studio / One API / local gateway)
-# AI_PROVIDER=openai_compatible
+# AI_PROVIDER=openai
+# AI_PROVIDER_API=openai-completions
 # AI_BASE_URL=http://127.0.0.1:8001/v1
 # AI_API_KEY=local-key
+# AI_MODEL=your-local-model
 
-# Claude-compatible example (Anthropic / MiniMax / Claude-compatible gateway)
-# AI_PROVIDER=claude_compatible
-# AI_BASE_URL=http://127.0.0.1:8002
+# Anthropic-compatible example (Claude-compatible gateway)
+# AI_PROVIDER=anthropic
+# AI_PROVIDER_API=anthropic-messages
+# AI_BASE_URL=http://127.0.0.1:8002/anthropic
 # AI_API_KEY=local-key
+# AI_MODEL=your-model
 # AI_MAX_TOKENS=1200
 # AI_ANTHROPIC_VERSION=2023-06-01
 
 # Ollama native example
-AI_BASE_URL=http://127.0.0.1:11434
-AI_API_KEY=
-AI_MAX_TOKENS=1200
-AI_ANTHROPIC_VERSION=2023-06-01
+# AI_PROVIDER=ollama
+# AI_PROVIDER_API=ollama-generate
+# AI_BASE_URL=http://127.0.0.1:11434
+# AI_API_KEY=
+# AI_MODEL=qwen2.5:7b

aiprovider/README.md

@@ -4,21 +4,31 @@
 
 完整使用说明见:
 
-- [docs/aiprovider.md](/home/ray/dev/linkong/planet/docs/aiprovider.md)
+- [docs/agents/aiprovider.md](/home/ray/dev/linkong/planet/docs/agents/aiprovider.md)
 
 当前支持:
 
-- `AI_PROVIDER=openai`
-- `AI_PROVIDER=openai_compatible`
-- `AI_PROVIDER=anthropic`
-- `AI_PROVIDER=anthropic_compatible`
-- `AI_PROVIDER=claude_compatible`
-- `AI_PROVIDER=ollama`
+- provider identity:
+  - `AI_PROVIDER=openai`
+  - `AI_PROVIDER=anthropic`
+  - `AI_PROVIDER=minimax`
+  - `AI_PROVIDER=ollama`
+- request adapter:
+  - `AI_PROVIDER_API=openai-completions`
+  - `AI_PROVIDER_API=anthropic-messages`
+  - `AI_PROVIDER_API=ollama-generate`
+
+兼容别名仍然保留:
+
+- `openai_compatible`
+- `anthropic_compatible`
+- `claude_compatible`
 
 典型配置:
 
 ```env
-AI_PROVIDER=openai_compatible
+AI_PROVIDER=openai
+AI_PROVIDER_API=openai-completions
 AI_BASE_URL=https://api.openai.com/v1
 AI_API_KEY=your_api_key
 AI_MODEL=gpt-4o-mini
@@ -26,13 +36,14 @@ AI_TIMEOUT_SECONDS=60
 AI_PROVIDER_SERVICE_TOKEN=change_me
 ```
 
-Claude 兼容供应商示例:
+MiniMax 中国大陆节点示例:
 
 ```env
-AI_PROVIDER=claude_compatible
-AI_BASE_URL=https://your-claude-compatible-endpoint.example.com
-AI_API_KEY=your_api_key
-AI_MODEL=your-claude-compatible-model
+AI_PROVIDER=minimax
+AI_PROVIDER_API=anthropic-messages
+AI_BASE_URL=https://api.minimaxi.com/anthropic
+AI_API_KEY=sk-cp-xxxxx
+AI_MODEL=MiniMax-M2.7
 AI_TIMEOUT_SECONDS=60
 AI_MAX_TOKENS=1200
 AI_ANTHROPIC_VERSION=2023-06-01
@@ -43,12 +54,15 @@ AI_PROVIDER_SERVICE_TOKEN=change_me
 
 - Anthropic 官方 Claude API
 - Claude 兼容网关
-- MiniMax 等提供 Claude/Anthropic 风格消息接口的服务
+- MiniMax 等提供 Anthropic Messages 风格接口的服务
+
+这套命名方式参考了 OpenClaw 的接入模式: provider 负责标识供应商, `AI_PROVIDER_API` 负责标识协议适配层, 避免把“供应商”和“协议”绑死在一起。
 
 Ollama 原生示例:
 
 ```env
 AI_PROVIDER=ollama
+AI_PROVIDER_API=ollama-generate
 AI_BASE_URL=http://127.0.0.1:11434
 AI_API_KEY=
 AI_MODEL=qwen2.5:7b
@@ -58,8 +72,8 @@ AI_PROVIDER_SERVICE_TOKEN=change_me
 
 本地模型接入建议:
 
-- `vLLM`、`LM Studio`、`One API`：优先使用 `openai_compatible`
-- `MiniMax`、Claude 兼容网关：使用 `claude_compatible`
+- `vLLM`、`LM Studio`、`One API`：`AI_PROVIDER=openai` + `AI_PROVIDER_API=openai-completions`
+- `MiniMax`、Claude 兼容网关：`AI_PROVIDER=minimax|anthropic` + `AI_PROVIDER_API=anthropic-messages`
 - `Ollama`：可直接使用 `ollama`
 
 启动模板:

aiprovider/config.py

@@ -9,6 +9,7 @@ class Settings(BaseSettings):
     SERVICE_VERSION: str = "0.1.0"
 
     AI_PROVIDER: str = "disabled"
+    AI_PROVIDER_API: str = "auto"
     AI_BASE_URL: str = "https://api.openai.com/v1"
     AI_API_KEY: str = ""
     AI_MODEL: str = ""

aiprovider/provider_service.py

@@ -8,6 +8,7 @@ from fastapi import HTTPException, status
 
 from aiprovider.config import settings
 from aiprovider.schemas import (
+    AIContentBlock,
     AIProviderStatusResponse,
     SituationalAnalysisRequest,
     SituationalAnalysisResponse,
@@ -18,9 +19,39 @@ def _normalize_provider(value: str) -> str:
     return (value or "disabled").strip().lower()
 
 
+def _normalize_provider_api(value: str) -> str:
+    return (value or "auto").strip().lower().replace("_", "-")
+
+
+def _resolve_provider_api(provider: str, configured_api: str) -> str:
+    if configured_api and configured_api != "auto":
+        return configured_api
+
+    if provider in {"openai", "openai-compatible", "openai_compatible"}:
+        return "openai-completions"
+    if provider in {
+        "anthropic",
+        "anthropic-compatible",
+        "anthropic_compatible",
+        "claude-compatible",
+        "claude_compatible",
+        "minimax",
+        "kimi-coding",
+        "moonshot-anthropic",
+    }:
+        return "anthropic-messages"
+    if provider == "ollama":
+        return "ollama-generate"
+    return "disabled"
+
+
 class ProviderService:
     def __init__(self) -> None:
         self.provider = _normalize_provider(settings.AI_PROVIDER)
+        self.provider_api = _resolve_provider_api(
+            self.provider,
+            _normalize_provider_api(settings.AI_PROVIDER_API),
+        )
         self.base_url = settings.AI_BASE_URL.rstrip("/")
         self.api_key = settings.AI_API_KEY
         self.default_model = settings.AI_MODEL
@@ -32,9 +63,11 @@ class ProviderService:
 
     def get_status(self) -> AIProviderStatusResponse:
         enabled = self.provider != "disabled"
-        configured = enabled and bool(self.base_url and self.api_key and self.default_model)
+        has_credentials = bool(self.api_key) if self._requires_api_key() else True
+        configured = enabled and bool(self.base_url and has_credentials and self.default_model)
         return AIProviderStatusResponse(
             provider=self.provider,
+            api=self.provider_api if enabled else None,
             enabled=enabled,
             configured=configured,
             model=self.default_model or None,
@@ -49,7 +82,8 @@ class ProviderService:
             )
 
         model = payload.preferred_model or self.default_model
-        if not self.base_url or not self.api_key or not model:
+        has_credentials = bool(self.api_key) if self._requires_api_key() else True
+        if not self.base_url or not has_credentials or not model:
             raise HTTPException(
                 status_code=status.HTTP_503_SERVICE_UNAVAILABLE,
                 detail="AI provider is not fully configured. Check AI_BASE_URL, AI_API_KEY, and AI_MODEL.",
@@ -57,28 +91,40 @@ class ProviderService:
 
         prompt = self._build_prompt(payload)
 
-        if self.provider in {"openai", "openai_compatible"}:
+        if self.provider_api == "openai-completions":
             data = await self._request_openai_compatible(model, prompt)
             content = self._extract_openai_content(data)
-        elif self.provider in {"anthropic", "anthropic_compatible", "claude_compatible"}:
-            data = await self._request_anthropic_compatible(model, prompt)
+            content_blocks = self._extract_openai_blocks(data)
+        elif self.provider_api == "anthropic-messages":
+            data = await self._request_anthropic_messages(model, prompt, payload.thinking)
             content = self._extract_anthropic_content(data)
-        elif self.provider == "ollama":
+            content_blocks = self._extract_anthropic_blocks(data)
+        elif self.provider_api == "ollama-generate":
             data = await self._request_ollama(model, prompt)
             content = self._extract_ollama_content(data)
+            content_blocks = self._extract_ollama_blocks(data)
         else:
             raise HTTPException(
                 status_code=status.HTTP_400_BAD_REQUEST,
-                detail=f"Unsupported AI provider: {self.provider}",
+                detail=f"Unsupported AI provider API: {self.provider_api}",
             )
 
+        text_blocks = [block.text for block in content_blocks if block.text]
+        thinking_blocks = [block.thinking for block in content_blocks if block.thinking]
+
         return SituationalAnalysisResponse(
             provider=self.provider,
             model=model,
             content=content,
+            content_blocks=content_blocks,
+            text_blocks=text_blocks,
+            thinking_blocks=thinking_blocks,
             raw_response=data,
         )
 
+    def _requires_api_key(self) -> bool:
+        return self.provider_api != "ollama-generate"
+
     def _build_prompt(self, payload: SituationalAnalysisRequest) -> str:
         sections = [
             f"任务标题:\n{payload.title}",
@@ -113,7 +159,12 @@ class ProviderService:
             request_body=request_body,
         )
 
-    async def _request_anthropic_compatible(self, model: str, prompt: str) -> dict[str, Any]:
+    async def _request_anthropic_messages(
+        self,
+        model: str,
+        prompt: str,
+        thinking: dict[str, Any] | None = None,
+    ) -> dict[str, Any]:
         request_body = {
             "model": model,
             "system": self.system_prompt,
@@ -131,8 +182,15 @@ class ProviderService:
             "max_tokens": self.max_tokens,
             "temperature": 0.2,
         }
+        resolved_thinking = self._resolve_anthropic_thinking(thinking)
+        if resolved_thinking:
+            request_body["thinking"] = resolved_thinking
+        if self.provider == "minimax" and self.base_url.endswith("/anthropic"):
+            path = "/v1/messages"
+        else:
+            path = "/messages"
         return await self._post(
-            path="/messages",
+            path=path,
             headers={
                 "x-api-key": self.api_key,
                 "anthropic-version": self.anthropic_version,
@@ -141,6 +199,25 @@ class ProviderService:
             request_body=request_body,
         )
 
+    def _resolve_anthropic_thinking(self, thinking: dict[str, Any] | None) -> dict[str, Any] | None:
+        if thinking:
+            return thinking
+
+        # OpenClaw treats MiniMax's Anthropic-compatible path specially:
+        # disable thinking by default unless the caller explicitly opts in.
+        if self.provider == "minimax":
+            return {"type": "disabled"}
+
+        return None
+
+    async def _request_anthropic_compatible(
+        self,
+        model: str,
+        prompt: str,
+        thinking: dict[str, Any] | None = None,
+    ) -> dict[str, Any]:
+        return await self._request_anthropic_messages(model, prompt, thinking)
+
     async def _request_ollama(self, model: str, prompt: str) -> dict[str, Any]:
         request_body = {
             "model": model,
@@ -218,6 +295,30 @@ class ProviderService:
             )
         return ""
 
+    def _extract_openai_blocks(self, payload: dict[str, Any]) -> list[AIContentBlock]:
+        choices = payload.get("choices") or []
+        if not choices:
+            return []
+
+        message = choices[0].get("message") or {}
+        content = message.get("content")
+        if isinstance(content, str):
+            return [AIContentBlock(type="text", text=content)]
+        if not isinstance(content, list):
+            return []
+
+        blocks: list[AIContentBlock] = []
+        for item in content:
+            if not isinstance(item, dict):
+                continue
+            blocks.append(
+                AIContentBlock(
+                    type=str(item.get("type", "text")),
+                    text=item.get("text") if isinstance(item.get("text"), str) else None,
+                    metadata={k: v for k, v in item.items() if k not in {"type", "text"}},
+                )
+            )
+        return blocks
     def _extract_anthropic_content(self, payload: dict[str, Any]) -> str:
         content = payload.get("content")
         if isinstance(content, str):
@@ -233,8 +334,39 @@ class ProviderService:
                 fragments.append(item["text"])
         return "".join(fragments)
 
+    def _extract_anthropic_blocks(self, payload: dict[str, Any]) -> list[AIContentBlock]:
+        content = payload.get("content")
+        if not isinstance(content, list):
+            return []
+
+        blocks: list[AIContentBlock] = []
+        for item in content:
+            if not isinstance(item, dict):
+                continue
+            blocks.append(
+                AIContentBlock(
+                    type=str(item.get("type", "unknown")),
+                    text=item.get("text") if isinstance(item.get("text"), str) else None,
+                    thinking=item.get("thinking") if isinstance(item.get("thinking"), str) else None,
+                    signature=item.get("signature") if isinstance(item.get("signature"), str) else None,
+                    metadata={
+                        k: v
+                        for k, v in item.items()
+                        if k not in {"type", "text", "thinking", "signature"}
+                    },
+                )
+            )
+        return blocks
+
+
     def _extract_ollama_content(self, payload: dict[str, Any]) -> str:
         response = payload.get("response")
         if isinstance(response, str):
             return response
         return ""
+
+    def _extract_ollama_blocks(self, payload: dict[str, Any]) -> list[AIContentBlock]:
+        response = payload.get("response")
+        if isinstance(response, str) and response:
+            return [AIContentBlock(type="text", text=response)]
+        return []

aiprovider/schemas.py

@@ -3,6 +3,14 @@ from typing import Any
 from pydantic import BaseModel, Field
 
 
+class AIContentBlock(BaseModel):
+    type: str
+    text: str | None = None
+    thinking: str | None = None
+    signature: str | None = None
+    metadata: dict[str, Any] = Field(default_factory=dict)
+
+
 class SituationalAnalysisRequest(BaseModel):
     title: str = Field(..., min_length=1, max_length=200)
     objective: str = Field(..., min_length=1, max_length=1000)
@@ -10,17 +18,22 @@ class SituationalAnalysisRequest(BaseModel):
     observations: list[str] = Field(default_factory=list)
     constraints: list[str] = Field(default_factory=list)
     preferred_model: str | None = Field(default=None, max_length=200)
+    thinking: dict[str, Any] | None = None
 
 
 class SituationalAnalysisResponse(BaseModel):
     provider: str
     model: str
     content: str
+    content_blocks: list[AIContentBlock] = Field(default_factory=list)
+    text_blocks: list[str] = Field(default_factory=list)
+    thinking_blocks: list[str] = Field(default_factory=list)
     raw_response: dict[str, Any] = Field(default_factory=dict)
 
 
 class AIProviderStatusResponse(BaseModel):
     provider: str
+    api: str | None = None
     enabled: bool
     configured: bool
     model: str | None = None

backend/app/api/main.py

@@ -13,7 +13,9 @@ from app.api.v1 import (
     collected_data,
     visualization,
     bgp,
+    news,
     system_control,
+    tv,
 )
 
 api_router = APIRouter()
@@ -33,3 +35,5 @@ api_router.include_router(settings.router, prefix="/settings", tags=["settings"]
 api_router.include_router(system_control.router, prefix="/system", tags=["system"])
 api_router.include_router(visualization.router, prefix="/visualization", tags=["visualization"])
 api_router.include_router(bgp.router, prefix="/bgp", tags=["bgp"])
+api_router.include_router(tv.router, prefix="/tv", tags=["tv"])
+api_router.include_router(news.router, prefix="/news", tags=["news"])

backend/app/api/v1/ai.py

@@ -1,15 +1,52 @@
 from uuid import uuid4
 
-from fastapi import APIRouter, Depends, Request, Response
+from fastapi import APIRouter, Depends, HTTPException, Request, Response
+from sqlalchemy.ext.asyncio import AsyncSession
 
 from app.core.security import get_current_user
+from app.db.session import get_db
 from app.models.user import User
 from app.schemas.ai import (
     AIProviderStatusResponse,
+    AlertBriefRequest,
+    AlertBriefResponse,
+    BGPBriefRequest,
+    BGPBriefRecordResponse,
+    BGPBriefRecordSummary,
+    PlaygroundMessageActionResponse,
+    PlaygroundMessageCreateRequest,
+    PlaygroundMessageEditRequest,
+    PlaygroundMessageResendRequest,
+    PlaygroundMessageStopRequest,
+    PlaygroundSessionResponse,
+    PlaygroundSessionUpsertRequest,
+    PlaygroundThreadResponse,
+    SituationalAlertBriefRequest,
+    SituationalAlertBriefResponse,
     SituationalAnalysisRequest,
     SituationalAnalysisResponse,
 )
+from app.services.alert_ai_brief import build_alert_brief_request
 from app.services.ai_client import AIProviderClient, get_ai_provider_client
+from app.services.bgp_ai_brief import build_bgp_brief_request
+from app.services.bgp_ai_brief_store import (
+    get_bgp_brief_record,
+    get_latest_bgp_brief_record,
+    list_bgp_brief_records,
+    save_bgp_brief_record,
+)
+from app.services.playground_session_store import (
+    get_playground_session,
+    upsert_playground_session,
+)
+from app.services.playground_chat_service import (
+    create_turn,
+    edit_user_message,
+    get_thread,
+    resend_turn,
+    stop_message,
+)
+from app.services.situational_alert_ai_brief import build_situational_alert_brief_request
 
 router = APIRouter()
 
@@ -37,3 +74,208 @@ async def analyze_situational_awareness(
     request_id = request.headers.get("X-Request-ID") or str(uuid4())
     response.headers["X-Request-ID"] = request_id
     return await provider_client.analyze(payload, request_id=request_id)
+
+
+@router.get("/playground/thread", response_model=PlaygroundThreadResponse | None)
+async def get_playground_thread(
+    session_key: str = "default",
+    current_user: User = Depends(get_current_user),
+    db: AsyncSession = Depends(get_db),
+):
+    return await get_thread(
+        db,
+        user_id=current_user.id,
+        session_key=session_key,
+    )
+
+
+@router.get("/playground/session", response_model=PlaygroundSessionResponse | None)
+async def get_saved_playground_session(
+    session_key: str = "default",
+    current_user: User = Depends(get_current_user),
+    db: AsyncSession = Depends(get_db),
+):
+    return await get_playground_session(
+        db,
+        user_id=current_user.id,
+        session_key=session_key,
+    )
+
+
+@router.put("/playground/session", response_model=PlaygroundSessionResponse)
+async def save_playground_session(
+    payload: PlaygroundSessionUpsertRequest,
+    current_user: User = Depends(get_current_user),
+    db: AsyncSession = Depends(get_db),
+):
+    return await upsert_playground_session(
+        db,
+        user_id=current_user.id,
+        payload=payload,
+    )
+
+
+@router.post("/playground/messages", response_model=PlaygroundMessageActionResponse)
+async def create_playground_message(
+    payload: PlaygroundMessageCreateRequest,
+    current_user: User = Depends(get_current_user),
+    provider_client: AIProviderClient = Depends(get_ai_provider_client),
+    db: AsyncSession = Depends(get_db),
+):
+    return await create_turn(
+        db,
+        user_id=current_user.id,
+        payload=payload,
+        provider_client=provider_client,
+    )
+
+
+@router.post("/playground/messages/stop", response_model=PlaygroundMessageActionResponse)
+async def stop_playground_message(
+    payload: PlaygroundMessageStopRequest,
+    current_user: User = Depends(get_current_user),
+    db: AsyncSession = Depends(get_db),
+):
+    return await stop_message(
+        db,
+        user_id=current_user.id,
+        payload=payload,
+    )
+
+
+@router.post("/playground/messages/resend", response_model=PlaygroundMessageActionResponse)
+async def resend_playground_message(
+    payload: PlaygroundMessageResendRequest,
+    current_user: User = Depends(get_current_user),
+    provider_client: AIProviderClient = Depends(get_ai_provider_client),
+    db: AsyncSession = Depends(get_db),
+):
+    return await resend_turn(
+        db,
+        user_id=current_user.id,
+        payload=payload,
+        provider_client=provider_client,
+    )
+
+
+@router.post("/playground/messages/edit", response_model=PlaygroundMessageActionResponse)
+async def edit_playground_message(
+    payload: PlaygroundMessageEditRequest,
+    current_user: User = Depends(get_current_user),
+    db: AsyncSession = Depends(get_db),
+):
+    return await edit_user_message(
+        db,
+        user_id=current_user.id,
+        payload=payload,
+    )
+
+
+@router.get("/bgp/briefs", response_model=list[BGPBriefRecordSummary])
+async def list_saved_bgp_briefs(
+    current_user: User = Depends(get_current_user),
+):
+    return list_bgp_brief_records()
+
+
+@router.get("/bgp/briefs/latest", response_model=BGPBriefRecordResponse | None)
+async def get_latest_saved_bgp_brief(
+    current_user: User = Depends(get_current_user),
+):
+    return get_latest_bgp_brief_record()
+
+
+@router.get("/bgp/briefs/{brief_id}", response_model=BGPBriefRecordResponse)
+async def get_saved_bgp_brief(
+    brief_id: str,
+    current_user: User = Depends(get_current_user),
+):
+    record = get_bgp_brief_record(brief_id)
+    if record is None:
+        raise HTTPException(status_code=404, detail="BGP brief not found")
+    return record
+
+
+@router.post("/bgp/brief", response_model=BGPBriefRecordResponse)
+async def analyze_bgp_brief(
+    payload: BGPBriefRequest,
+    request: Request,
+    response: Response,
+    current_user: User = Depends(get_current_user),
+    provider_client: AIProviderClient = Depends(get_ai_provider_client),
+    db: AsyncSession = Depends(get_db),
+):
+    request_id = request.headers.get("X-Request-ID") or str(uuid4())
+    response.headers["X-Request-ID"] = request_id
+
+    brief_request, facts, context = await build_bgp_brief_request(
+        db,
+        incident_limit=payload.incident_limit,
+        anomaly_limit=payload.anomaly_limit,
+        collector_limit=payload.collector_limit,
+    )
+    brief_request.preferred_model = payload.preferred_model
+    brief_request.thinking = payload.thinking
+
+    analysis = await provider_client.analyze(brief_request, request_id=request_id)
+    return save_bgp_brief_record(
+        analysis,
+        request_id=request_id,
+        facts=facts,
+        context=context,
+    )
+
+
+@router.post("/alerts/brief", response_model=AlertBriefResponse)
+async def analyze_alert_brief(
+    payload: AlertBriefRequest,
+    request: Request,
+    response: Response,
+    current_user: User = Depends(get_current_user),
+    provider_client: AIProviderClient = Depends(get_ai_provider_client),
+    db: AsyncSession = Depends(get_db),
+):
+    request_id = request.headers.get("X-Request-ID") or str(uuid4())
+    response.headers["X-Request-ID"] = request_id
+
+    brief_request, facts, context = await build_alert_brief_request(
+        db,
+        alert_limit=payload.alert_limit,
+    )
+    brief_request.preferred_model = payload.preferred_model
+    brief_request.thinking = payload.thinking
+
+    analysis = await provider_client.analyze(brief_request, request_id=request_id)
+    return AlertBriefResponse(
+        **analysis.model_dump(),
+        title=brief_request.title,
+        objective=brief_request.objective,
+        facts=facts,
+        context=context,
+    )
+
+
+@router.post("/situational-alerts/brief", response_model=SituationalAlertBriefResponse)
+async def analyze_situational_alert_brief(
+    payload: SituationalAlertBriefRequest,
+    request: Request,
+    response: Response,
+    current_user: User = Depends(get_current_user),
+    provider_client: AIProviderClient = Depends(get_ai_provider_client),
+    db: AsyncSession = Depends(get_db),
+):
+    request_id = request.headers.get("X-Request-ID") or str(uuid4())
+    response.headers["X-Request-ID"] = request_id
+
+    brief_request, facts, context = await build_situational_alert_brief_request(db)
+    brief_request.preferred_model = payload.preferred_model
+    brief_request.thinking = payload.thinking
+
+    analysis = await provider_client.analyze(brief_request, request_id=request_id)
+    return SituationalAlertBriefResponse(
+        **analysis.model_dump(),
+        title=brief_request.title,
+        objective=brief_request.objective,
+        facts=facts,
+        context=context,
+    )

backend/app/api/v1/alerts.py

@@ -1,7 +1,7 @@
 from datetime import UTC, datetime
 from typing import Optional
 
-from fastapi import APIRouter, Depends
+from fastapi import APIRouter, Depends, HTTPException
 from sqlalchemy import select, func, case
 from sqlalchemy.ext.asyncio import AsyncSession
 
@@ -9,6 +9,7 @@ from app.db.session import get_db
 from app.models.user import User
 from app.core.security import get_current_user
 from app.models.alert import Alert, AlertSeverity, AlertStatus
+from app.schemas.alert import AlertResolutionRequest
 
 
 router = APIRouter()
@@ -77,7 +78,7 @@ async def acknowledge_alert(
 @router.post("/{alert_id}/resolve")
 async def resolve_alert(
     alert_id: int,
-    resolution: str,
+    payload: AlertResolutionRequest,
     current_user: User = Depends(get_current_user),
     db: AsyncSession = Depends(get_db),
 ):
@@ -85,12 +86,12 @@ async def resolve_alert(
     alert = result.scalar_one_or_none()
 
     if not alert:
-        return {"error": "Alert not found"}
+        raise HTTPException(status_code=404, detail="Alert not found")
 
     alert.status = AlertStatus.RESOLVED
     alert.resolved_by = current_user.id
     alert.resolved_at = datetime.now(UTC)
-    alert.resolution_notes = resolution
+    alert.resolution_notes = payload.resolution
     await db.commit()
 
     return {"message": "Alert resolved", "alert": alert.to_dict()}
@@ -101,25 +102,44 @@ async def get_alert_stats(
     current_user: User = Depends(get_current_user),
     db: AsyncSession = Depends(get_db),
 ):
-    critical_query = select(func.count(Alert.id)).where(
-        Alert.severity == AlertSeverity.CRITICAL,
-        Alert.status == AlertStatus.ACTIVE,
+    result = await db.execute(
+        select(
+            func.sum(
+                case(
+                    (
+                        (Alert.severity == AlertSeverity.CRITICAL)
+                        & (Alert.status == AlertStatus.ACTIVE),
+                        1,
+                    ),
+                    else_=0,
+                )
+            ).label("critical"),
+            func.sum(
+                case(
+                    (
+                        (Alert.severity == AlertSeverity.WARNING)
+                        & (Alert.status == AlertStatus.ACTIVE),
+                        1,
+                    ),
+                    else_=0,
+                )
+            ).label("warning"),
+            func.sum(
+                case(
+                    (
+                        (Alert.severity == AlertSeverity.INFO)
+                        & (Alert.status == AlertStatus.ACTIVE),
+                        1,
+                    ),
+                    else_=0,
+                )
+            ).label("info"),
+        )
     )
-    warning_query = select(func.count(Alert.id)).where(
-        Alert.severity == AlertSeverity.WARNING,
-        Alert.status == AlertStatus.ACTIVE,
-    )
-    info_query = select(func.count(Alert.id)).where(
-        Alert.severity == AlertSeverity.INFO,
-        Alert.status == AlertStatus.ACTIVE,
-    )
-
-    critical_result = await db.execute(critical_query)
-    warning_result = await db.execute(warning_query)
-    info_result = await db.execute(info_query)
+    row = result.one()
 
     return {
-        "critical": critical_result.scalar() or 0,
-        "warning": warning_result.scalar() or 0,
-        "info": info_result.scalar() or 0,
+        "critical": row.critical or 0,
+        "warning": row.warning or 0,
+        "info": row.info or 0,
     }

backend/app/api/v1/bgp.py

@@ -22,16 +22,161 @@ def _parse_dt(value: Optional[str]) -> Optional[datetime]:
     if not value:
         return None
     return datetime.fromisoformat(value.replace("Z", "+00:00"))
+def _event_filters(
+    *,
+    prefix: Optional[str],
+    origin_asn: Optional[int],
+    peer_asn: Optional[int],
+    collector: Optional[str],
+    event_type: Optional[str],
+    source: Optional[str],
+    time_from: Optional[datetime],
+    time_to: Optional[datetime],
+):
+    filters = [BGPObservation.source.in_(BGP_SOURCES)]
+    if source:
+        filters.append(BGPObservation.source == source)
+    if prefix:
+        filters.append(BGPObservation.prefix == prefix)
+    if origin_asn is not None:
+        filters.append(BGPObservation.origin_asn == origin_asn)
+    if peer_asn is not None:
+        filters.append(BGPObservation.peer_asn == peer_asn)
+    if collector:
+        filters.append(BGPObservation.collector == collector)
+    if event_type:
+        filters.append(BGPObservation.event_type == event_type)
+    if time_from:
+        filters.append(BGPObservation.observed_at >= time_from)
+    if time_to:
+        filters.append(BGPObservation.observed_at <= time_to)
+    return filters
 
 
-def _matches_time(value: Optional[datetime], time_from: Optional[datetime], time_to: Optional[datetime]) -> bool:
-    if value is None:
-        return False
-    if time_from and value < time_from:
-        return False
-    if time_to and value > time_to:
-        return False
-    return True
+def _anomaly_filters(
+    *,
+    severity: Optional[str],
+    anomaly_type: Optional[str],
+    status: Optional[str],
+    prefix: Optional[str],
+    origin_asn: Optional[int],
+    time_from: Optional[datetime],
+    time_to: Optional[datetime],
+):
+    filters = []
+    if severity:
+        filters.append(BGPAnomaly.severity == severity)
+    if anomaly_type:
+        filters.append(BGPAnomaly.anomaly_type == anomaly_type)
+    if status:
+        filters.append(BGPAnomaly.status == status)
+    if prefix:
+        filters.append(BGPAnomaly.prefix == prefix)
+    if origin_asn is not None:
+        filters.append(BGPAnomaly.origin_asn == origin_asn)
+    if time_from:
+        filters.append(BGPAnomaly.created_at >= time_from)
+    if time_to:
+        filters.append(BGPAnomaly.created_at <= time_to)
+    return filters
+
+
+def _incident_filters(
+    *,
+    severity: Optional[str],
+    incident_type: Optional[str],
+    status: Optional[str],
+):
+    filters = []
+    if severity:
+        filters.append(BGPIncident.severity == severity)
+    if incident_type:
+        filters.append(BGPIncident.incident_type == incident_type)
+    if status:
+        filters.append(BGPIncident.status == status)
+    return filters
+
+
+async def _build_event_summary_payload(db: AsyncSession) -> dict:
+    base_filters = [BGPObservation.source.in_(BGP_SOURCES)]
+
+    total_result = await db.execute(
+        select(func.count(BGPObservation.id)).where(*base_filters)
+    )
+    collectors_result = await db.execute(
+        select(func.count(func.distinct(BGPObservation.collector))).where(
+            *base_filters, BGPObservation.collector.isnot(None)
+        )
+    )
+    prefixes_result = await db.execute(
+        select(func.count(func.distinct(BGPObservation.prefix))).where(
+            *base_filters, BGPObservation.prefix.isnot(None)
+        )
+    )
+    type_result = await db.execute(
+        select(BGPObservation.event_type, func.count(BGPObservation.id))
+        .where(*base_filters)
+        .group_by(BGPObservation.event_type)
+    )
+
+    return {
+        "total": total_result.scalar() or 0,
+        "collector_count": collectors_result.scalar() or 0,
+        "prefix_count": prefixes_result.scalar() or 0,
+        "by_type": {row[0]: row[1] for row in type_result.fetchall()},
+    }
+
+
+async def _build_anomaly_summary_payload(db: AsyncSession) -> dict:
+    total_result = await db.execute(select(func.count(BGPAnomaly.id)))
+    type_result = await db.execute(
+        select(BGPAnomaly.anomaly_type, func.count(BGPAnomaly.id))
+        .group_by(BGPAnomaly.anomaly_type)
+        .order_by(func.count(BGPAnomaly.id).desc())
+    )
+    severity_result = await db.execute(
+        select(BGPAnomaly.severity, func.count(BGPAnomaly.id))
+        .group_by(BGPAnomaly.severity)
+        .order_by(func.count(BGPAnomaly.id).desc())
+    )
+    status_result = await db.execute(
+        select(BGPAnomaly.status, func.count(BGPAnomaly.id))
+        .group_by(BGPAnomaly.status)
+        .order_by(func.count(BGPAnomaly.id).desc())
+    )
+
+    return {
+        "total": total_result.scalar() or 0,
+        "by_type": {row[0]: row[1] for row in type_result.fetchall()},
+        "by_severity": {row[0]: row[1] for row in severity_result.fetchall()},
+        "by_status": {row[0]: row[1] for row in status_result.fetchall()},
+    }
+
+
+async def _build_incident_summary_payload(db: AsyncSession) -> dict:
+    total_result = await db.execute(select(func.count(BGPIncident.id)))
+    type_result = await db.execute(
+        select(BGPIncident.incident_type, func.count(BGPIncident.id))
+        .group_by(BGPIncident.incident_type)
+        .order_by(func.count(BGPIncident.id).desc())
+    )
+    severity_result = await db.execute(
+        select(BGPIncident.severity, func.count(BGPIncident.id))
+        .group_by(BGPIncident.severity)
+        .order_by(func.count(BGPIncident.id).desc())
+    )
+    status_result = await db.execute(
+        select(BGPIncident.status, func.count(BGPIncident.id))
+        .group_by(BGPIncident.status)
+        .order_by(func.count(BGPIncident.id).desc())
+    )
+
+    return {
+        "total": total_result.scalar() or 0,
+        "by_type": {row[0]: row[1] for row in type_result.fetchall()},
+        "by_severity": {row[0]: row[1] for row in severity_result.fetchall()},
+        "by_status": {row[0]: row[1] for row in status_result.fetchall()},
+    }
 
 
 @router.get("/events")
@@ -49,41 +194,36 @@ async def list_bgp_events(
     current_user: User = Depends(get_current_user),
     db: AsyncSession = Depends(get_db),
 ):
-    stmt = (
-        select(BGPObservation)
-        .where(BGPObservation.source.in_(BGP_SOURCES))
-        .order_by(BGPObservation.observed_at.desc(), BGPObservation.id.desc())
-    )
-    if source:
-        stmt = stmt.where(BGPObservation.source == source)
-
-    result = await db.execute(stmt)
-    records = result.scalars().all()
     dt_from = _parse_dt(time_from)
     dt_to = _parse_dt(time_to)
-
-    filtered = []
-    for record in records:
-        if prefix and record.prefix != prefix:
-            continue
-        if origin_asn is not None and record.origin_asn != origin_asn:
-            continue
-        if peer_asn is not None and record.peer_asn != peer_asn:
-            continue
-        if collector and record.collector != collector:
-            continue
-        if event_type and record.event_type != event_type:
-            continue
-        if (dt_from or dt_to) and not _matches_time(record.observed_at, dt_from, dt_to):
-            continue
-        filtered.append(record)
-
+    filters = _event_filters(
+        prefix=prefix,
+        origin_asn=origin_asn,
+        peer_asn=peer_asn,
+        collector=collector,
+        event_type=event_type,
+        source=source,
+        time_from=dt_from,
+        time_to=dt_to,
+    )
     offset = (page - 1) * page_size
+    count_result = await db.execute(
+        select(func.count(BGPObservation.id)).where(*filters)
+    )
+    data_result = await db.execute(
+        select(BGPObservation)
+        .where(*filters)
+        .order_by(BGPObservation.observed_at.desc(), BGPObservation.id.desc())
+        .offset(offset)
+        .limit(page_size)
+    )
+    records = data_result.scalars().all()
+
     return {
-        "total": len(filtered),
+        "total": count_result.scalar() or 0,
         "page": page,
         "page_size": page_size,
-        "data": [record.to_dict() for record in filtered[offset : offset + page_size]],
+        "data": [record.to_dict() for record in records],
     }
 
 
@@ -92,21 +232,7 @@ async def get_bgp_event_summary(
     current_user: User = Depends(get_current_user),
     db: AsyncSession = Depends(get_db),
 ):
-    result = await db.execute(select(BGPObservation).where(BGPObservation.source.in_(BGP_SOURCES)))
-    records = result.scalars().all()
-
-    collectors = sorted({record.collector for record in records if record.collector})
-    prefixes = sorted({record.prefix for record in records if record.prefix})
-    by_type: dict[str, int] = {}
-    for record in records:
-        by_type[record.event_type] = by_type.get(record.event_type, 0) + 1
-
-    return {
-        "total": len(records),
-        "collector_count": len(collectors),
-        "prefix_count": len(prefixes),
-        "by_type": by_type,
-    }
+    return await _build_event_summary_payload(db)
 
 
 @router.get("/collectors")
@@ -138,6 +264,32 @@ async def get_bgp_collector_summary(
     }
 
 
+@router.get("/overview/summary")
+async def get_bgp_overview_summary(
+    current_user: User = Depends(get_current_user),
+    db: AsyncSession = Depends(get_db),
+):
+    event_summary = await _build_event_summary_payload(db)
+    anomaly_summary = await _build_anomaly_summary_payload(db)
+    incident_summary = await _build_incident_summary_payload(db)
+    collectors = await build_bgp_collector_coverage(db, source_filter=BGP_SOURCES)
+    active_collectors = [item for item in collectors if item["observation_count"] > 0]
+
+    return {
+        "incidentSummary": incident_summary,
+        "anomalySummary": anomaly_summary,
+        "eventSummary": event_summary,
+        "collectorSummary": {
+            "total": len(collectors),
+            "active_collectors": len(active_collectors),
+            "observed_prefixes": sum(item["prefix_count"] for item in active_collectors),
+            "observed_origins": sum(item["origin_asn_count"] for item in active_collectors),
+            "recent_24h_events": sum(item["recent_24h_observation_count"] for item in active_collectors),
+            "recent_7d_events": sum(item["recent_7d_observation_count"] for item in active_collectors),
+        },
+    }
+
+
 @router.get("/events/{event_id}")
 async def get_bgp_event(
     event_id: int,
@@ -164,31 +316,35 @@ async def list_bgp_anomalies(
     current_user: User = Depends(get_current_user),
     db: AsyncSession = Depends(get_db),
 ):
-    stmt = select(BGPAnomaly).order_by(BGPAnomaly.created_at.desc(), BGPAnomaly.id.desc())
-    if severity:
-        stmt = stmt.where(BGPAnomaly.severity == severity)
-    if anomaly_type:
-        stmt = stmt.where(BGPAnomaly.anomaly_type == anomaly_type)
-    if status:
-        stmt = stmt.where(BGPAnomaly.status == status)
-    if prefix:
-        stmt = stmt.where(BGPAnomaly.prefix == prefix)
-    if origin_asn is not None:
-        stmt = stmt.where(BGPAnomaly.origin_asn == origin_asn)
-
-    result = await db.execute(stmt)
-    records = result.scalars().all()
     dt_from = _parse_dt(time_from)
     dt_to = _parse_dt(time_to)
-    if dt_from or dt_to:
-        records = [record for record in records if _matches_time(record.created_at, dt_from, dt_to)]
-
+    filters = _anomaly_filters(
+        severity=severity,
+        anomaly_type=anomaly_type,
+        status=status,
+        prefix=prefix,
+        origin_asn=origin_asn,
+        time_from=dt_from,
+        time_to=dt_to,
+    )
     offset = (page - 1) * page_size
+    total_result = await db.execute(
+        select(func.count(BGPAnomaly.id)).where(*filters)
+    )
+    data_result = await db.execute(
+        select(BGPAnomaly)
+        .where(*filters)
+        .order_by(BGPAnomaly.created_at.desc(), BGPAnomaly.id.desc())
+        .offset(offset)
+        .limit(page_size)
+    )
+    records = data_result.scalars().all()
+
     return {
-        "total": len(records),
+        "total": total_result.scalar() or 0,
         "page": page,
         "page_size": page_size,
-        "data": [record.to_dict() for record in records[offset : offset + page_size]],
+        "data": [record.to_dict() for record in records],
     }
 
 
@@ -197,29 +353,7 @@ async def get_bgp_anomaly_summary(
     current_user: User = Depends(get_current_user),
     db: AsyncSession = Depends(get_db),
 ):
-    total_result = await db.execute(select(func.count(BGPAnomaly.id)))
-    type_result = await db.execute(
-        select(BGPAnomaly.anomaly_type, func.count(BGPAnomaly.id))
-        .group_by(BGPAnomaly.anomaly_type)
-        .order_by(func.count(BGPAnomaly.id).desc())
-    )
-    severity_result = await db.execute(
-        select(BGPAnomaly.severity, func.count(BGPAnomaly.id))
-        .group_by(BGPAnomaly.severity)
-        .order_by(func.count(BGPAnomaly.id).desc())
-    )
-    status_result = await db.execute(
-        select(BGPAnomaly.status, func.count(BGPAnomaly.id))
-        .group_by(BGPAnomaly.status)
-        .order_by(func.count(BGPAnomaly.id).desc())
-    )
-
-    return {
-        "total": total_result.scalar() or 0,
-        "by_type": {row[0]: row[1] for row in type_result.fetchall()},
-        "by_severity": {row[0]: row[1] for row in severity_result.fetchall()},
-        "by_status": {row[0]: row[1] for row in status_result.fetchall()},
-    }
+    return await _build_anomaly_summary_payload(db)
 
 
 @router.get("/anomalies/{anomaly_id}")
@@ -244,22 +378,29 @@ async def list_bgp_incidents(
     current_user: User = Depends(get_current_user),
     db: AsyncSession = Depends(get_db),
 ):
-    stmt = select(BGPIncident).order_by(BGPIncident.created_at.desc(), BGPIncident.id.desc())
-    if severity:
-        stmt = stmt.where(BGPIncident.severity == severity)
-    if incident_type:
-        stmt = stmt.where(BGPIncident.incident_type == incident_type)
-    if status:
-        stmt = stmt.where(BGPIncident.status == status)
-
-    result = await db.execute(stmt)
-    records = result.scalars().all()
+    filters = _incident_filters(
+        severity=severity,
+        incident_type=incident_type,
+        status=status,
+    )
     offset = (page - 1) * page_size
+    total_result = await db.execute(
+        select(func.count(BGPIncident.id)).where(*filters)
+    )
+    data_result = await db.execute(
+        select(BGPIncident)
+        .where(*filters)
+        .order_by(BGPIncident.created_at.desc(), BGPIncident.id.desc())
+        .offset(offset)
+        .limit(page_size)
+    )
+    records = data_result.scalars().all()
+
     return {
-        "total": len(records),
+        "total": total_result.scalar() or 0,
         "page": page,
         "page_size": page_size,
-        "data": [record.to_dict() for record in records[offset : offset + page_size]],
+        "data": [record.to_dict() for record in records],
     }
 
 
@@ -268,29 +409,7 @@ async def get_bgp_incident_summary(
     current_user: User = Depends(get_current_user),
     db: AsyncSession = Depends(get_db),
 ):
-    total_result = await db.execute(select(func.count(BGPIncident.id)))
-    type_result = await db.execute(
-        select(BGPIncident.incident_type, func.count(BGPIncident.id))
-        .group_by(BGPIncident.incident_type)
-        .order_by(func.count(BGPIncident.id).desc())
-    )
-    severity_result = await db.execute(
-        select(BGPIncident.severity, func.count(BGPIncident.id))
-        .group_by(BGPIncident.severity)
-        .order_by(func.count(BGPIncident.id).desc())
-    )
-    status_result = await db.execute(
-        select(BGPIncident.status, func.count(BGPIncident.id))
-        .group_by(BGPIncident.status)
-        .order_by(func.count(BGPIncident.id).desc())
-    )
-
-    return {
-        "total": total_result.scalar() or 0,
-        "by_type": {row[0]: row[1] for row in type_result.fetchall()},
-        "by_severity": {row[0]: row[1] for row in severity_result.fetchall()},
-        "by_status": {row[0]: row[1] for row in status_result.fetchall()},
-    }
+    return await _build_incident_summary_payload(db)
 
 
 @router.get("/incidents/{incident_id}")

backend/app/api/v1/dashboard.py

@@ -2,7 +2,7 @@
 
 from datetime import UTC, datetime, timedelta
 from fastapi import APIRouter, Depends
-from sqlalchemy import select, func, text
+from sqlalchemy import case, select, func, text
 from sqlalchemy.ext.asyncio import AsyncSession
 
 from app.db.session import get_db
@@ -118,58 +118,77 @@ async def get_stats(
     built_in_count = len(COLLECTOR_INFO)
     built_in_active = built_in_count  # Built-in are always "active" for counting purposes
 
-    # Count custom configs from database
-    result = await db.execute(select(func.count(DataSourceConfig.id)))
-    custom_count = result.scalar() or 0
-
     result = await db.execute(
-        select(func.count(DataSourceConfig.id)).where(DataSourceConfig.is_active == True)
+        select(
+            func.count(DataSourceConfig.id).label("custom_count"),
+            func.sum(
+                case((DataSourceConfig.is_active == True, 1), else_=0)
+            ).label("custom_active"),
+        )
     )
-    custom_active = result.scalar() or 0
+    datasource_stats = result.one()
+    custom_count = datasource_stats.custom_count or 0
+    custom_active = datasource_stats.custom_active or 0
 
     # Total datasources
     total_datasources = built_in_count + custom_count
     active_datasources = built_in_active + custom_active
 
-    # Tasks today (from database)
     result = await db.execute(
-        select(func.count(CollectionTask.id)).where(CollectionTask.started_at >= today_start)
-    )
-    tasks_today = result.scalar() or 0
-
-    result = await db.execute(
-        select(func.count(CollectionTask.id)).where(
-            CollectionTask.status == "success",
-            CollectionTask.started_at >= today_start,
+        select(
+            func.count(CollectionTask.id).label("tasks_today"),
+            func.sum(
+                case(
+                    (CollectionTask.status == "success", 1),
+                    else_=0,
+                )
+            ).label("success_tasks"),
         )
+        .where(CollectionTask.started_at >= today_start)
     )
-    success_tasks = result.scalar() or 0
+    task_stats = result.one()
+    tasks_today = task_stats.tasks_today or 0
+    success_tasks = task_stats.success_tasks or 0
     success_rate = (success_tasks / tasks_today * 100) if tasks_today > 0 else 0
 
-    # Alerts
     result = await db.execute(
-        select(func.count(Alert.id)).where(
-            Alert.severity == AlertSeverity.CRITICAL,
-            Alert.status == "active",
+        select(
+            func.sum(
+                case(
+                    (
+                        (Alert.severity == AlertSeverity.CRITICAL)
+                        & (Alert.status == "active"),
+                        1,
+                    ),
+                    else_=0,
+                )
+            ).label("critical_alerts"),
+            func.sum(
+                case(
+                    (
+                        (Alert.severity == AlertSeverity.WARNING)
+                        & (Alert.status == "active"),
+                        1,
+                    ),
+                    else_=0,
+                )
+            ).label("warning_alerts"),
+            func.sum(
+                case(
+                    (
+                        (Alert.severity == AlertSeverity.INFO)
+                        & (Alert.status == "active"),
+                        1,
+                    ),
+                    else_=0,
+                )
+            ).label("info_alerts"),
         )
     )
-    critical_alerts = result.scalar() or 0
-
-    result = await db.execute(
-        select(func.count(Alert.id)).where(
-            Alert.severity == AlertSeverity.WARNING,
-            Alert.status == "active",
-        )
-    )
-    warning_alerts = result.scalar() or 0
-
-    result = await db.execute(
-        select(func.count(Alert.id)).where(
-            Alert.severity == AlertSeverity.INFO,
-            Alert.status == "active",
-        )
-    )
-    info_alerts = result.scalar() or 0
+    alert_stats = result.one()
+    critical_alerts = alert_stats.critical_alerts or 0
+    warning_alerts = alert_stats.warning_alerts or 0
+    info_alerts = alert_stats.info_alerts or 0
 
     response = {
         "total_datasources": total_datasources,

backend/app/api/v1/datasources.py

@@ -3,7 +3,7 @@ from datetime import datetime, timedelta, timezone
 from typing import Optional
 
 from fastapi import APIRouter, Depends, HTTPException, Query
-from sqlalchemy import func, select
+from sqlalchemy import func, select, text
 from sqlalchemy.ext.asyncio import AsyncSession
 
 from app.core.time import to_iso8601_utc
@@ -11,10 +11,17 @@ from app.core.security import get_current_user
 from app.core.data_sources import get_data_sources_config
 from app.db.session import get_db
 from app.models.collected_data import CollectedData
+from app.models.data_snapshot import DataSnapshot
 from app.models.datasource import DataSource
+from app.models.datasource_config import DataSourceConfig
 from app.models.task import CollectionTask
 from app.models.user import User
-from app.services.scheduler import get_latest_task_id_for_datasource, run_collector_now, sync_datasource_job
+from app.services.scheduler import (
+    cancel_running_collector_now,
+    get_latest_task_id_for_datasource,
+    run_collector_now,
+    sync_datasource_job,
+)
 
 router = APIRouter()
 STALE_RUNNING_TASK_TIMEOUT_MINUTES = 90
@@ -34,6 +41,156 @@ def is_due_for_collection(datasource: DataSource, now: datetime) -> bool:
     return datasource.last_run_at + timedelta(minutes=datasource.frequency_minutes) <= now
 
 
+def _task_rank_column(order_column):
+    return func.row_number().over(
+        partition_by=CollectionTask.datasource_id,
+        order_by=(order_column.desc().nullslast(), CollectionTask.id.desc()),
+    ).label("row_num")
+
+
+async def _load_latest_running_tasks(
+    db: AsyncSession,
+    datasource_ids: list[int],
+) -> dict[int, CollectionTask]:
+    if not datasource_ids:
+        return {}
+
+    ranked_tasks = (
+        select(
+            CollectionTask.id.label("task_id"),
+            _task_rank_column(CollectionTask.started_at),
+        )
+        .where(CollectionTask.datasource_id.in_(datasource_ids))
+        .where(CollectionTask.status == "running")
+        .subquery()
+    )
+    result = await db.execute(
+        select(CollectionTask)
+        .join(ranked_tasks, CollectionTask.id == ranked_tasks.c.task_id)
+        .where(ranked_tasks.c.row_num == 1)
+    )
+    return {task.datasource_id: task for task in result.scalars().all()}
+
+
+async def _load_latest_completed_tasks(
+    db: AsyncSession,
+    datasource_ids: list[int],
+) -> dict[int, CollectionTask]:
+    if not datasource_ids:
+        return {}
+
+    ranked_tasks = (
+        select(
+            CollectionTask.id.label("task_id"),
+            _task_rank_column(CollectionTask.completed_at),
+        )
+        .where(CollectionTask.datasource_id.in_(datasource_ids))
+        .where(CollectionTask.completed_at.isnot(None))
+        .where(CollectionTask.status.in_(("success", "failed", "cancelled")))
+        .subquery()
+    )
+    result = await db.execute(
+        select(CollectionTask)
+        .join(ranked_tasks, CollectionTask.id == ranked_tasks.c.task_id)
+        .where(ranked_tasks.c.row_num == 1)
+    )
+    return {task.datasource_id: task for task in result.scalars().all()}
+
+
+async def _load_latest_task_ids(
+    db: AsyncSession,
+    datasource_ids: list[int],
+) -> dict[int, int]:
+    if not datasource_ids:
+        return {}
+
+    ranked_tasks = (
+        select(
+            CollectionTask.id.label("task_id"),
+            CollectionTask.datasource_id.label("datasource_id"),
+            func.row_number().over(
+                partition_by=CollectionTask.datasource_id,
+                order_by=CollectionTask.id.desc(),
+            ).label("row_num"),
+        )
+        .where(CollectionTask.datasource_id.in_(datasource_ids))
+        .subquery()
+    )
+    result = await db.execute(
+        select(ranked_tasks.c.datasource_id, ranked_tasks.c.task_id)
+        .where(ranked_tasks.c.row_num == 1)
+    )
+    return {datasource_id: task_id for datasource_id, task_id in result.all()}
+
+
+async def _load_datasource_data_counts(
+    db: AsyncSession,
+    sources: list[str],
+) -> dict[str, int]:
+    if not sources:
+        return {}
+
+    result = await db.execute(
+        select(CollectedData.source, func.count(CollectedData.id))
+        .where(CollectedData.source.in_(sources))
+        .group_by(CollectedData.source)
+    )
+    return {source: count for source, count in result.all()}
+
+
+async def _load_datasource_endpoint_overrides(
+    db: AsyncSession,
+    sources: list[str],
+) -> dict[str, str]:
+    if not sources:
+        return {}
+
+    result = await db.execute(
+        select(DataSourceConfig.name, DataSourceConfig.endpoint)
+        .where(DataSourceConfig.name.in_(sources))
+        .where(DataSourceConfig.is_active.is_(True))
+        .where(DataSourceConfig.endpoint.isnot(None))
+    )
+    return {
+        name: endpoint
+        for name, endpoint in result.all()
+        if endpoint
+    }
+
+
+async def _load_datasource_list_context(
+    db: AsyncSession,
+    datasources: list[DataSource],
+) -> tuple[dict[int, CollectionTask], dict[int, CollectionTask], dict[str, int], dict[str, str]]:
+    datasource_ids = [datasource.id for datasource in datasources]
+    sources = [datasource.source for datasource in datasources]
+
+    running_tasks = await _load_latest_running_tasks(db, datasource_ids)
+    datasource_by_id = {datasource.id: datasource for datasource in datasources}
+    now = datetime.now(timezone.utc)
+
+    stale_datasource_ids: list[int] = []
+    for datasource_id, task in running_tasks.items():
+        started_at = task.started_at
+        if started_at is None:
+            continue
+        if started_at.tzinfo is None:
+            started_at = started_at.replace(tzinfo=timezone.utc)
+        if now - started_at > timedelta(minutes=STALE_RUNNING_TASK_TIMEOUT_MINUTES):
+            datasource = datasource_by_id.get(datasource_id)
+            if datasource is not None:
+                await fail_and_rollback_stale_running_task(db, datasource, task)
+                stale_datasource_ids.append(datasource_id)
+
+    if stale_datasource_ids:
+        running_tasks = await _load_latest_running_tasks(db, datasource_ids)
+
+    completed_tasks = await _load_latest_completed_tasks(db, datasource_ids)
+    data_counts = await _load_datasource_data_counts(db, sources)
+    endpoint_overrides = await _load_datasource_endpoint_overrides(db, sources)
+    return running_tasks, completed_tasks, data_counts, endpoint_overrides
+
+
 async def get_datasource_record(db: AsyncSession, source_id: str) -> Optional[DataSource]:
     datasource = None
     try:
@@ -52,18 +209,6 @@ async def get_datasource_record(db: AsyncSession, source_id: str) -> Optional[Da
     return result.scalar_one_or_none()
 
 
-async def get_last_completed_task(db: AsyncSession, datasource_id: int) -> Optional[CollectionTask]:
-    result = await db.execute(
-        select(CollectionTask)
-        .where(CollectionTask.datasource_id == datasource_id)
-        .where(CollectionTask.completed_at.isnot(None))
-        .where(CollectionTask.status.in_(("success", "failed", "cancelled")))
-        .order_by(CollectionTask.completed_at.desc())
-        .limit(1)
-    )
-    return result.scalar_one_or_none()
-
-
 async def get_running_task(db: AsyncSession, datasource_id: int) -> Optional[CollectionTask]:
     result = await db.execute(
         select(CollectionTask)
@@ -87,17 +232,152 @@ async def get_running_task(db: AsyncSession, datasource_id: int) -> Optional[Col
     if now - started_at <= timedelta(minutes=STALE_RUNNING_TASK_TIMEOUT_MINUTES):
         return task
 
-    existing_error = (task.error_message or "").strip()
+    datasource = await db.get(DataSource, datasource_id)
+    if datasource is not None:
+        await fail_and_rollback_stale_running_task(db, datasource, task)
+    else:
+        existing_error = (task.error_message or "").strip()
+        stale_reason = (
+            f"Marked failed automatically after stale running timeout "
+            f"({STALE_RUNNING_TASK_TIMEOUT_MINUTES}m)"
+        )
+        task.status = "failed"
+        task.phase = "failed"
+        task.completed_at = now
+        task.error_message = f"{existing_error}\n{stale_reason}".strip() if existing_error else stale_reason
+        await db.commit()
+    return None
+
+
+async def rollback_orphaned_running_task(
+    db: AsyncSession,
+    datasource: DataSource,
+    running_task: CollectionTask,
+) -> None:
+    snapshot_result = await db.execute(
+        select(DataSnapshot)
+        .where(
+            DataSnapshot.datasource_id == datasource.id,
+            DataSnapshot.task_id == running_task.id,
+        )
+        .order_by(DataSnapshot.id.desc())
+        .limit(1)
+    )
+    snapshot = snapshot_result.scalar_one_or_none()
+
+    await db.execute(CollectedData.__table__.delete().where(CollectedData.task_id == running_task.id))
+
+    await db.execute(
+        text(
+            """
+            UPDATE collected_data
+            SET is_current = FALSE
+            WHERE source = :source
+            """
+        ),
+        {"source": datasource.source},
+    )
+
+    if snapshot is not None:
+        snapshot.status = "cancelled"
+        snapshot.is_current = False
+        snapshot.completed_at = datetime.now(timezone.utc)
+        summary = dict(snapshot.summary or {})
+        summary["rollback"] = True
+        summary["rollback_reason"] = "orphaned_running_task_after_backend_restart"
+        snapshot.summary = summary
+
+        if snapshot.parent_snapshot_id is not None:
+            parent_snapshot = await db.get(DataSnapshot, snapshot.parent_snapshot_id)
+            if parent_snapshot:
+                parent_snapshot.is_current = True
+            await db.execute(
+                text(
+                    """
+                    UPDATE collected_data
+                    SET is_current = TRUE
+                    WHERE snapshot_id = :snapshot_id
+                    """
+                ),
+                {"snapshot_id": snapshot.parent_snapshot_id},
+            )
+
+    running_task.status = "cancelled"
+    running_task.phase = "cancelled"
+    running_task.completed_at = datetime.now(timezone.utc)
+    existing_error = (running_task.error_message or "").strip()
+    cancel_reason = "Cancelled after backend restart because the running task handle was lost; incomplete writes rolled back"
+    running_task.error_message = f"{existing_error}\n{cancel_reason}".strip() if existing_error else cancel_reason
+    datasource.last_status = "cancelled"
+    datasource.last_run_at = datetime.now(timezone.utc)
+    await db.commit()
+
+
+async def fail_and_rollback_stale_running_task(
+    db: AsyncSession,
+    datasource: DataSource,
+    running_task: CollectionTask,
+) -> None:
+    snapshot_result = await db.execute(
+        select(DataSnapshot)
+        .where(
+            DataSnapshot.datasource_id == datasource.id,
+            DataSnapshot.task_id == running_task.id,
+        )
+        .order_by(DataSnapshot.id.desc())
+        .limit(1)
+    )
+    snapshot = snapshot_result.scalar_one_or_none()
+
+    await db.execute(CollectedData.__table__.delete().where(CollectedData.task_id == running_task.id))
+
+    await db.execute(
+        text(
+            """
+            UPDATE collected_data
+            SET is_current = FALSE
+            WHERE source = :source
+            """
+        ),
+        {"source": datasource.source},
+    )
+
+    if snapshot is not None:
+        snapshot.status = "failed"
+        snapshot.is_current = False
+        snapshot.completed_at = datetime.now(timezone.utc)
+        summary = dict(snapshot.summary or {})
+        summary["rollback"] = True
+        summary["rollback_reason"] = "stale_running_task_timeout"
+        snapshot.summary = summary
+
+        if snapshot.parent_snapshot_id is not None:
+            parent_snapshot = await db.get(DataSnapshot, snapshot.parent_snapshot_id)
+            if parent_snapshot:
+                parent_snapshot.is_current = True
+            await db.execute(
+                text(
+                    """
+                    UPDATE collected_data
+                    SET is_current = TRUE
+                    WHERE snapshot_id = :snapshot_id
+                    """
+                ),
+                {"snapshot_id": snapshot.parent_snapshot_id},
+            )
+
+    existing_error = (running_task.error_message or "").strip()
     stale_reason = (
         f"Marked failed automatically after stale running timeout "
-        f"({STALE_RUNNING_TASK_TIMEOUT_MINUTES}m)"
+        f"({STALE_RUNNING_TASK_TIMEOUT_MINUTES}m); incomplete writes rolled back"
     )
-    task.status = "failed"
-    task.phase = "failed"
-    task.completed_at = now
-    task.error_message = f"{existing_error}\n{stale_reason}".strip() if existing_error else stale_reason
+    running_task.status = "failed"
+    running_task.phase = "failed"
+    running_task.completed_at = datetime.now(timezone.utc)
+    running_task.error_message = f"{existing_error}\n{stale_reason}".strip() if existing_error else stale_reason
+    datasource.last_status = "failed"
+    datasource.last_run_at = datetime.now(timezone.utc)
     await db.commit()
-    return None
 
 
 @router.get("")
@@ -121,14 +401,17 @@ async def list_datasources(
 
     collector_list = []
     config = get_data_sources_config()
+    running_tasks, completed_tasks, data_counts, endpoint_overrides = await _load_datasource_list_context(
+        db,
+        datasources,
+    )
     for datasource in datasources:
-        running_task = await get_running_task(db, datasource.id)
-        last_task = await get_last_completed_task(db, datasource.id)
-        endpoint = await config.get_url(datasource.source, db)
-        data_count_result = await db.execute(
-            select(func.count(CollectedData.id)).where(CollectedData.source == datasource.source)
+        running_task = running_tasks.get(datasource.id)
+        last_task = completed_tasks.get(datasource.id)
+        endpoint = endpoint_overrides.get(datasource.source) or config.get_yaml_url(
+            datasource.source,
         )
-        data_count = data_count_result.scalar() or 0
+        data_count = data_counts.get(datasource.source, 0)
 
         last_run_at = datasource.last_run_at or (last_task.completed_at if last_task else None)
         last_run = to_iso8601_utc(last_run_at)
@@ -137,6 +420,7 @@ async def list_datasources(
         collector_list.append(
             {
                 "id": datasource.id,
+                "source": datasource.source,
                 "name": datasource.name,
                 "module": datasource.module,
                 "priority": datasource.priority,
@@ -189,9 +473,13 @@ async def trigger_all_datasources(
     skipped_sources: list[dict] = []
     failed_sources: list[dict] = []
     now = datetime.now(timezone.utc)
+    running_tasks = await _load_latest_running_tasks(
+        db,
+        [datasource.id for datasource in datasources],
+    )
 
     for datasource in datasources:
-        running_task = await get_running_task(db, datasource.id)
+        running_task = running_tasks.get(datasource.id)
         if running_task is not None:
             skipped_sources.append(
                 {
@@ -219,7 +507,7 @@ async def trigger_all_datasources(
             )
             continue
 
-        previous_task_ids[datasource.id] = await get_latest_task_id_for_datasource(datasource.id)
+        previous_task_ids[datasource.id] = None
         success = run_collector_now(datasource.source)
         if not success:
             failed_sources.append(
@@ -241,13 +529,24 @@ async def trigger_all_datasources(
             }
         )
 
+    latest_task_ids = await _load_latest_task_ids(
+        db,
+        [datasource.id for datasource in datasources],
+    )
+    for datasource_id in previous_task_ids:
+        previous_task_ids[datasource_id] = latest_task_ids.get(datasource_id)
+
     for _ in range(20):
         await asyncio.sleep(0.1)
         pending = [item for item in triggered_sources if item["task_id"] is None]
         if not pending:
             break
+        latest_task_ids = await _load_latest_task_ids(
+            db,
+            [item["id"] for item in pending],
+        )
         for item in pending:
-            task_id = await get_latest_task_id_for_datasource(item["id"])
+            task_id = latest_task_ids.get(item["id"])
             if task_id is not None and task_id != previous_task_ids.get(item["id"]):
                 item["task_id"] = task_id
 
@@ -346,6 +645,7 @@ async def get_datasource_stats(
 @router.post("/{source_id}/trigger")
 async def trigger_datasource(
     source_id: str,
+    force: bool = Query(False),
     current_user: User = Depends(get_current_user),
     db: AsyncSession = Depends(get_db),
 ):
@@ -356,6 +656,26 @@ async def trigger_datasource(
     if not datasource.is_active:
         raise HTTPException(status_code=400, detail="Data source is disabled")
 
+    running_task = await get_running_task(db, datasource.id)
+    if running_task is not None and not force:
+        raise HTTPException(
+            status_code=409,
+            detail={
+                "reason": "running_task_in_progress",
+                "message": "当前采集任务尚未完成，重新触发会丢失本次未完成进度。是否强制重新采集？",
+                "task_id": running_task.id,
+                "phase": running_task.phase,
+                "progress": running_task.progress,
+                "records_processed": running_task.records_processed,
+                "total_records": running_task.total_records,
+            },
+        )
+
+    if running_task is not None and force:
+        cancelled = await cancel_running_collector_now(datasource.source)
+        if not cancelled:
+            await rollback_orphaned_running_task(db, datasource, running_task)
+
     previous_task_id = await get_latest_task_id_for_datasource(datasource.id)
     success = run_collector_now(datasource.source)
     if not success:
@@ -375,6 +695,7 @@ async def trigger_datasource(
         "source_id": datasource.id,
         "task_id": task_id,
         "collector_name": datasource.source,
+        "force": force,
         "message": f"Collector '{datasource.source}' has been triggered",
     }
 
@@ -412,6 +733,7 @@ async def clear_datasource_data(
 async def get_task_status(
     source_id: str,
     task_id: Optional[int] = None,
+    current_user: User = Depends(get_current_user),
     db: AsyncSession = Depends(get_db),
 ):
     datasource = await get_datasource_record(db, source_id)

backend/app/api/v1/news.py

@@ -0,0 +1,13 @@
+from fastapi import APIRouter, Query
+
+from app.services.earth_news import get_earth_news_payload
+
+router = APIRouter()
+
+
+@router.get("/earth-feed")
+async def get_earth_feed(
+    lat: float | None = Query(None, description="Current Earth view center latitude"),
+    lon: float | None = Query(None, description="Current Earth view center longitude"),
+):
+    return await get_earth_news_payload(lat=lat, lon=lon)

backend/app/api/v1/settings.py

@@ -1,3 +1,4 @@
+from copy import deepcopy
 from datetime import UTC, datetime
 from typing import Optional
 
@@ -13,6 +14,7 @@ from app.models.datasource import DataSource
 from app.models.system_setting import SystemSetting
 from app.models.user import User
 from app.services.scheduler import sync_datasource_job
+from app.services.tv_streams import DEFAULT_TV_SETTINGS, get_tv_settings_payload, normalize_tv_settings
 
 router = APIRouter()
 
@@ -36,6 +38,7 @@ DEFAULT_SETTINGS = {
         "max_login_attempts": 5,
         "password_policy": "medium",
     },
+    "tv": DEFAULT_TV_SETTINGS,
 }
 
 
@@ -67,8 +70,34 @@ class CollectorSettingsUpdate(BaseModel):
     frequency_minutes: int = Field(default=60, ge=1, le=10080)
 
 
+class TVStreamSourceUpdate(BaseModel):
+    id: str = Field(min_length=1, max_length=100)
+    name: str = Field(min_length=1, max_length=200)
+    provider: str = Field(default="Unknown", max_length=100)
+    region: str = Field(default="Global", max_length=100)
+    language: str = Field(default="und", max_length=32)
+    source_type: str = Field(default="iframe", pattern="^(iframe|hls|video|external|youtube)$")
+    embed_url: str = ""
+    stream_url: str = ""
+    homepage_url: str = ""
+    poster_url: str = ""
+    youtube_video_id: str = ""
+    youtube_channel: str = ""
+    is_enabled: bool = True
+    is_fallback: bool = False
+    sort_order: int = Field(default=10, ge=0, le=9999)
+    collector_source: Optional[str] = None
+    notes: str = ""
+
+
+class TVSettingsUpdate(BaseModel):
+    default_source_id: str = Field(default=DEFAULT_TV_SETTINGS["default_source_id"], min_length=1)
+    auto_fallback: bool = True
+    sources: list[TVStreamSourceUpdate] = Field(default_factory=list)
+
+
 def merge_with_defaults(category: str, payload: Optional[dict]) -> dict:
-    merged = DEFAULT_SETTINGS[category].copy()
+    merged = deepcopy(DEFAULT_SETTINGS[category])
     if payload:
         merged.update(payload)
     return merged
@@ -79,6 +108,26 @@ async def get_setting_record(db: AsyncSession, category: str) -> Optional[System
     return result.scalar_one_or_none()
 
 
+async def get_setting_payloads(db: AsyncSession, categories: list[str]) -> dict[str, dict]:
+    if not categories:
+        return {}
+
+    result = await db.execute(
+        select(SystemSetting).where(SystemSetting.category.in_(categories))
+    )
+    records_by_category = {
+        record.category: record
+        for record in result.scalars().all()
+    }
+    return {
+        category: merge_with_defaults(
+            category,
+            records_by_category.get(category).payload if records_by_category.get(category) else None,
+        )
+        for category in categories
+    }
+
+
 async def get_setting_payload(db: AsyncSession, category: str) -> dict:
     record = await get_setting_record(db, category)
     return merge_with_defaults(category, record.payload if record else None)
@@ -175,6 +224,25 @@ async def update_security_settings(
     return {"status": "updated", "security": payload}
 
 
+@router.get("/tv")
+async def get_tv_settings(
+    current_user: User = Depends(get_current_user),
+    db: AsyncSession = Depends(get_db),
+):
+    return {"tv": await get_tv_settings_payload(db)}
+
+
+@router.put("/tv")
+async def update_tv_settings(
+    settings: TVSettingsUpdate,
+    current_user: User = Depends(get_current_user),
+    db: AsyncSession = Depends(get_db),
+):
+    payload = normalize_tv_settings(settings.model_dump())
+    saved = await save_setting_payload(db, "tv", payload)
+    return {"status": "updated", "tv": normalize_tv_settings(saved)}
+
+
 @router.get("/collectors")
 async def get_collector_settings(
     current_user: User = Depends(get_current_user),
@@ -212,10 +280,15 @@ async def get_all_settings(
 ):
     result = await db.execute(select(DataSource).order_by(DataSource.module, DataSource.id))
     datasources = result.scalars().all()
+    setting_payloads = await get_setting_payloads(
+        db,
+        ["system", "notifications", "security"],
+    )
     return {
-        "system": await get_setting_payload(db, "system"),
-        "notifications": await get_setting_payload(db, "notifications"),
-        "security": await get_setting_payload(db, "security"),
+        "system": setting_payloads["system"],
+        "notifications": setting_payloads["notifications"],
+        "security": setting_payloads["security"],
+        "tv": await get_tv_settings_payload(db),
         "collectors": [serialize_collector(datasource) for datasource in datasources],
         "generated_at": to_iso8601_utc(datetime.now(UTC)),
     }

backend/app/api/v1/system_control.py

@@ -4,12 +4,15 @@ import os
 import subprocess
 import sys
 
-from fastapi import APIRouter, Depends, HTTPException, status
+from datetime import datetime
+
+from fastapi import APIRouter, Depends, HTTPException, Query, Request, status
 from pydantic import BaseModel
 
 from app.core.config import ROOT_DIR
 from app.core.security import get_current_user
 from app.models.user import User
+from app.services.persistent_logs import record_audit_log, record_system_log
 from app.services.system_control import (
     build_task_id,
     clear_active_task_id,
@@ -23,6 +26,15 @@ from app.services.system_control import (
     set_active_task_id,
     upsert_task_state,
 )
+from app.services.system_logs import (
+    DEFAULT_LOG_LINE_LIMIT,
+    MAX_LOG_LINE_LIMIT,
+    SUPPORTED_LOG_LEVELS,
+    append_buffer_log,
+    list_log_sources,
+    normalize_log_level,
+    read_log_snapshot,
+)
 
 router = APIRouter()
 
@@ -47,6 +59,59 @@ class RestartTaskLogsResponse(BaseModel):
     lines: list[str]
 
 
+class SystemLogSourceSummary(BaseModel):
+    source_id: str
+    name: str
+    kind: str
+    location: str
+    description: str
+    category: str
+    status: str
+
+
+class SystemLogSourcesResponse(BaseModel):
+    items: list[SystemLogSourceSummary]
+
+
+class SystemLogDailyMarker(BaseModel):
+    date_token: str
+    total: int
+    dominant_level: str
+
+
+class SystemLogSnapshotResponse(BaseModel):
+    source_id: str
+    name: str
+    kind: str
+    location: str
+    description: str
+    category: str
+    status: str
+    level: str
+    selected_levels: list[str] = []
+    search_query: str = ""
+    available_levels: list[str]
+    daily_markers: list[SystemLogDailyMarker] = []
+    line_limit: int
+    line_count: int
+    lines: list[str]
+
+
+class EarthClientLogEventCreate(BaseModel):
+    level: str = "error"
+    message: str
+    category: str | None = None
+    url: str | None = None
+    module: str | None = None
+    detail: str | None = None
+
+
+class EarthClientLogEventResponse(BaseModel):
+    accepted: bool
+    source_id: str
+    level: str
+
+
 def ensure_super_admin(current_user: User) -> None:
     if not require_super_admin(current_user.role):
         raise HTTPException(
@@ -55,9 +120,22 @@ def ensure_super_admin(current_user: User) -> None:
         )
 
 
+def validate_log_date(raw_value: str | None, field_name: str) -> str | None:
+    if raw_value in {None, ""}:
+        return None
+    try:
+        return datetime.strptime(raw_value, "%Y-%m-%d").date().isoformat()
+    except ValueError as exc:
+        raise HTTPException(
+            status_code=status.HTTP_400_BAD_REQUEST,
+            detail=f"{field_name} must be in YYYY-MM-DD format",
+        ) from exc
+
+
 @router.post("/restart-tasks", response_model=RestartTaskResponse)
 async def create_restart_task(
     payload: RestartTaskCreate,
+    request: Request,
     current_user: User = Depends(get_current_user),
 ):
     ensure_super_admin(current_user)
@@ -133,11 +211,31 @@ async def create_restart_task(
             requested_by=requested_by,
         )
         clear_active_task_id(task_id)
+        await record_audit_log(
+            action="system.restart_task.requested",
+            actor_id=current_user.id,
+            actor_name=current_user.username,
+            target_type="restart_task",
+            target_id=task_id,
+            result="failed",
+            ip=request.client.host if request.client else None,
+            details={"action": payload.action, "message": task_state["message"]},
+        )
         raise HTTPException(
             status_code=status.HTTP_500_INTERNAL_SERVER_ERROR,
             detail=task_state["message"],
         ) from exc
 
+    await record_audit_log(
+        action="system.restart_task.requested",
+        actor_id=current_user.id,
+        actor_name=current_user.username,
+        target_type="restart_task",
+        target_id=task_id,
+        result="accepted",
+        ip=request.client.host if request.client else None,
+        details={"action": payload.action},
+    )
     return task_state
 
 
@@ -165,3 +263,92 @@ async def get_restart_task_logs(
     if task is None:
         raise HTTPException(status_code=status.HTTP_404_NOT_FOUND, detail="Restart task not found")
     return {"task_id": task_id, "lines": get_task_logs(task_id)}
+
+
+@router.get("/logs/sources", response_model=SystemLogSourcesResponse)
+async def get_system_log_sources(
+    current_user: User = Depends(get_current_user),
+):
+    ensure_super_admin(current_user)
+    return {"items": list_log_sources()}
+
+
+@router.get("/logs/{source_id}", response_model=SystemLogSnapshotResponse)
+async def get_system_log_snapshot(
+    source_id: str,
+    limit: int = DEFAULT_LOG_LINE_LIMIT,
+    level: str = "all",
+    levels: str | None = Query(None, description="Comma-separated log levels"),
+    start_date: str | None = Query(None, description="Filter logs from this date (YYYY-MM-DD)"),
+    end_date: str | None = Query(None, description="Filter logs until this date (YYYY-MM-DD)"),
+    search: str | None = Query(None, description="Case-insensitive substring search"),
+    current_user: User = Depends(get_current_user),
+):
+    ensure_super_admin(current_user)
+
+    if limit < 1 or limit > MAX_LOG_LINE_LIMIT:
+        raise HTTPException(
+            status_code=status.HTTP_400_BAD_REQUEST,
+            detail=f"limit must be between 1 and {MAX_LOG_LINE_LIMIT}",
+        )
+    if str(level).strip().lower() not in SUPPORTED_LOG_LEVELS and normalize_log_level(level) == "all" and str(level).strip().lower() not in {"", "all"}:
+        raise HTTPException(status_code=status.HTTP_400_BAD_REQUEST, detail="Unsupported log level")
+    if levels:
+        for raw_level in str(levels).split(","):
+            normalized_level = str(raw_level).strip().lower()
+            if not normalized_level:
+                continue
+            if normalized_level not in SUPPORTED_LOG_LEVELS and normalize_log_level(normalized_level) == "all":
+                raise HTTPException(status_code=status.HTTP_400_BAD_REQUEST, detail="Unsupported log level")
+    normalized_start_date = validate_log_date(start_date, "start_date")
+    normalized_end_date = validate_log_date(end_date, "end_date")
+    if normalized_start_date and normalized_end_date and normalized_start_date > normalized_end_date:
+        raise HTTPException(status_code=status.HTTP_400_BAD_REQUEST, detail="start_date must be earlier than or equal to end_date")
+
+    snapshot = read_log_snapshot(
+        source_id,
+        limit,
+        level=level,
+        levels=levels,
+        start_date=normalized_start_date,
+        end_date=normalized_end_date,
+        search=search,
+    )
+    if snapshot is None:
+        raise HTTPException(status_code=status.HTTP_404_NOT_FOUND, detail="Log source not found")
+    return snapshot
+
+
+@router.post("/logs/earth-client", response_model=EarthClientLogEventResponse)
+async def ingest_earth_client_log(
+    payload: EarthClientLogEventCreate,
+    request: Request,
+):
+    normalized_level = normalize_log_level(payload.level)
+    append_buffer_log(
+        "earth-client",
+        level=normalized_level,
+        message=payload.message,
+        context={
+            "category": payload.category or "",
+            "url": payload.url or "",
+            "module": payload.module or "",
+            "detail": payload.detail or "",
+        },
+    )
+    await record_system_log(
+        source="earth-client",
+        service="earth",
+        module=payload.module or "earth-client",
+        event="earth.client.runtime_log",
+        level=normalized_level,
+        message=payload.message,
+        category=payload.category or "client-runtime",
+        context={
+            "url": payload.url or "",
+            "detail": payload.detail or "",
+            "module": payload.module or "",
+            "client_ip": request.client.host if request.client else "",
+        },
+    )
+    return {"accepted": True, "source_id": "earth-client", "level": normalized_level}

backend/app/api/v1/tv.py

@@ -0,0 +1,70 @@
+from urllib.parse import quote, urljoin
+
+import httpx
+from fastapi import APIRouter, Depends, HTTPException, Query
+from fastapi.responses import Response
+from sqlalchemy.ext.asyncio import AsyncSession
+
+from app.db.session import get_db
+from app.services.tv_streams import get_public_tv_payload, is_allowed_tv_proxy_url
+
+router = APIRouter()
+
+
+@router.get("/streams")
+async def list_public_tv_streams(
+    db: AsyncSession = Depends(get_db),
+):
+    return await get_public_tv_payload(db)
+
+
+@router.get("/proxy")
+async def proxy_tv_stream(
+    url: str = Query(..., description="Upstream TV stream or manifest URL"),
+    db: AsyncSession = Depends(get_db),
+):
+    payload = await get_public_tv_payload(db)
+    if not is_allowed_tv_proxy_url(url, payload.get("sources", [])):
+        raise HTTPException(status_code=403, detail="TV proxy target is not allowed")
+
+    try:
+        async with httpx.AsyncClient(follow_redirects=True, timeout=20.0) as client:
+            upstream = await client.get(
+                url,
+                headers={
+                    "User-Agent": "Mozilla/5.0",
+                    "Referer": "https://tv.cctv.com/live/cctv4/",
+                },
+            )
+            upstream.raise_for_status()
+    except httpx.HTTPError as exc:
+        raise HTTPException(status_code=502, detail=f"Failed to fetch TV stream: {exc}") from exc
+
+    content_type = upstream.headers.get("content-type", "application/octet-stream")
+    raw_content = upstream.content
+    response_url = str(upstream.url)
+    is_manifest = (
+        response_url.endswith(".m3u8")
+        or "mpegurl" in content_type.lower()
+        or raw_content.lstrip().startswith(b"#EXTM3U")
+    )
+
+    headers = {"Cache-Control": "no-store"}
+
+    if is_manifest:
+        manifest_text = upstream.text
+        rewritten_lines: list[str] = []
+        for line in manifest_text.splitlines():
+            stripped = line.strip()
+            if not stripped or stripped.startswith("#"):
+                rewritten_lines.append(line)
+                continue
+            absolute_url = urljoin(response_url, stripped)
+            rewritten_lines.append(f"/api/v1/tv/proxy?url={quote(absolute_url, safe='')}")
+        return Response(
+            content="\n".join(rewritten_lines),
+            media_type="application/vnd.apple.mpegurl",
+            headers=headers,
+        )
+
+    return Response(content=raw_content, media_type=content_type, headers=headers)

backend/app/api/v1/visualization.py

@@ -6,12 +6,14 @@ Returns GeoJSON format compatible with Three.js, CesiumJS, and Unreal Cesium.
 
 from datetime import UTC, datetime
 import math
-from fastapi import APIRouter, HTTPException, Depends, Query
+import httpx
+from fastapi import APIRouter, HTTPException, Depends, Query, Response
 from sqlalchemy.ext.asyncio import AsyncSession
 from sqlalchemy import select, func
 from typing import List, Dict, Any, Optional
 
 from app.core.collected_data_fields import get_record_field
+from app.core.countries import get_country_centroid
 from app.core.satellite_tle import build_tle_lines_from_elements
 from app.core.time import to_iso8601_utc
 from app.db.session import get_db
@@ -21,8 +23,14 @@ from app.models.collected_data import CollectedData
 from app.services.bgp_collectors import build_bgp_collector_coverage
 from app.services.cable_graph import build_graph_from_data, CableGraph, haversine_distance
 from app.services.collectors.bgp_common import RIPE_RIS_COLLECTOR_COORDS
+from app.services.persistent_logs import record_system_log
+from app.core.logging import get_logger
 
 router = APIRouter()
+logger = get_logger(__name__, service="api")
+TERRAIN_TILE_URL_TEMPLATE = (
+    "https://s3.amazonaws.com/elevation-tiles-prod/terrarium/{z}/{x}/{y}.png"
+)
 
 
 # ============== Converter Functions ==============
@@ -176,6 +184,12 @@ def convert_satellite_to_geojson(records: List[CollectedData]) -> Dict[str, Any]
                 mean_motion=metadata.get("mean_motion"),
             )
 
+        constellation_group = _normalize_satellite_constellation_group(
+            metadata.get("constellation_group"),
+            record.name,
+        )
+        footprint_policy = _get_satellite_footprint_policy(constellation_group)
+
         features.append(
             {
                 "type": "Feature",
@@ -185,6 +199,8 @@ def convert_satellite_to_geojson(records: List[CollectedData]) -> Dict[str, Any]
                     "id": record.id,
                     "norad_cat_id": norad_id,
                     "name": record.name,
+                    "constellation_group": constellation_group,
+                    "footprint_policy": footprint_policy,
                     "international_designator": metadata.get("international_designator"),
                     "epoch": metadata.get("epoch"),
                     "inclination": metadata.get("inclination"),
@@ -205,42 +221,109 @@ def convert_satellite_to_geojson(records: List[CollectedData]) -> Dict[str, Any]
     return {"type": "FeatureCollection", "features": features}
 
 
-def dedupe_satellite_records(records: List[CollectedData]) -> List[CollectedData]:
-    """Keep only the newest record for each satellite identity."""
-    latest_by_key: Dict[str, CollectedData] = {}
+def _normalize_satellite_constellation_group(
+    raw_group: Any,
+    name: Optional[str],
+) -> Optional[str]:
+    normalized_group = str(raw_group or "").strip().lower()
+    if normalized_group:
+        return normalized_group
 
-    for record in records:
-        metadata = record.extra_data or {}
-        norad_id = metadata.get("norad_cat_id")
-        dedupe_key = (
-            str(norad_id)
-            if norad_id not in (None, "")
-            else str(record.source_id or record.entity_key or record.name or record.id)
-        )
+    normalized_name = str(name or "").strip().upper()
+    if normalized_name.startswith("STARLINK"):
+        return "starlink"
+    if normalized_name.startswith("IRIDIUM"):
+        return "iridium-next"
 
-        existing = latest_by_key.get(dedupe_key)
-        if existing is None or (record.id or 0) > (existing.id or 0):
-            latest_by_key[dedupe_key] = record
-
-    return sorted(latest_by_key.values(), key=lambda item: item.id or 0, reverse=True)
+    return None
 
 
-def dedupe_collected_records(records: List[CollectedData]) -> List[CollectedData]:
-    """Keep only the newest record for each collected entity."""
-    latest_by_key: Dict[str, CollectedData] = {}
+def _get_satellite_footprint_policy(constellation_group: Optional[str]) -> str:
+    if constellation_group == "starlink":
+        return "starlink_ground_footprint"
+    if constellation_group == "iridium-next":
+        return "iridium_coverage_ring"
+    return "none"
 
-    for record in records:
-        dedupe_key = str(
-            record.source_id
-            or record.entity_key
-            or record.name
-            or record.id
-        )
-        existing = latest_by_key.get(dedupe_key)
-        if existing is None or (record.id or 0) > (existing.id or 0):
-            latest_by_key[dedupe_key] = record
 
-    return sorted(latest_by_key.values(), key=lambda item: item.id or 0, reverse=True)
+def _current_collected_data_stmt(source: str):
+    return (
+        select(CollectedData)
+        .where(CollectedData.source == source)
+        .where(CollectedData.is_current.is_(True))
+        .order_by(CollectedData.id.desc())
+    )
+
+
+async def _load_current_collected_data(
+    db: AsyncSession,
+    source: str,
+    *,
+    exclude_unknown_name: bool = False,
+    limit: Optional[int] = None,
+) -> List[CollectedData]:
+    stmt = _current_collected_data_stmt(source)
+    if exclude_unknown_name:
+        stmt = stmt.where(CollectedData.name != "Unknown")
+    if limit is not None:
+        stmt = stmt.limit(limit)
+
+    result = await db.execute(stmt)
+    return list(result.scalars().all())
+
+
+async def _load_current_collected_data_by_sources(
+    db: AsyncSession,
+    sources: List[str],
+) -> Dict[str, List[CollectedData]]:
+    if not sources:
+        return {}
+
+    stmt = (
+        select(CollectedData)
+        .where(CollectedData.source.in_(sources))
+        .where(CollectedData.is_current.is_(True))
+        .order_by(CollectedData.source.asc(), CollectedData.id.desc())
+    )
+    result = await db.execute(stmt)
+
+    grouped_records: Dict[str, List[CollectedData]] = {source: [] for source in sources}
+    for record in result.scalars().all():
+        grouped_records.setdefault(record.source, []).append(record)
+
+    return grouped_records
+
+
+def _build_landing_point_cable_maps(
+    relation_records: List[CollectedData],
+    cable_records: List[CollectedData],
+) -> tuple[Dict[int, List[int]], Dict[int, str]]:
+    city_to_cable_ids_map: Dict[int, List[int]] = {}
+    for relation_record in relation_records:
+        if not relation_record.extra_data:
+            continue
+        city_id = relation_record.extra_data.get("city_id")
+        cable_id = relation_record.extra_data.get("cable_id")
+        if city_id is None or cable_id is None:
+            continue
+        city_to_cable_ids_map.setdefault(city_id, [])
+        if cable_id not in city_to_cable_ids_map[city_id]:
+            city_to_cable_ids_map[city_id].append(cable_id)
+
+    cable_id_to_name_map: Dict[int, str] = {}
+    for cable_record in cable_records:
+        if not cable_record.extra_data:
+            continue
+        cable_id = cable_record.extra_data.get("cable_id")
+        cable_name = cable_record.name
+        if cable_id and cable_name:
+            cable_id_to_name_map[cable_id] = cable_name
+
+    return city_to_cable_ids_map, cable_id_to_name_map
+
+
+def _filter_known_records(records: List[CollectedData]) -> List[CollectedData]:
+    return [record for record in records if record.name != "Unknown"]
 
 
 def convert_supercomputer_to_geojson(records: List[CollectedData]) -> Dict[str, Any]:
@@ -317,6 +400,215 @@ def convert_gpu_cluster_to_geojson(records: List[CollectedData]) -> Dict[str, An
     return {"type": "FeatureCollection", "features": features}
 
 
+def _parse_float(value: Any) -> Optional[float]:
+    try:
+        if value in (None, ""):
+            return None
+        return float(value)
+    except (TypeError, ValueError):
+        return None
+
+
+COMPUTE_CENTER_COORDINATE_HINTS = (
+    ("el capitan", 37.6819, -121.7681),
+    ("livermore", 37.6819, -121.7681),
+    ("llnl", 37.6819, -121.7681),
+    ("lawrence livermore", 37.6819, -121.7681),
+    ("frontier", 35.9319, -84.3107),
+    ("oak ridge", 35.9319, -84.3107),
+    ("ornl", 35.9319, -84.3107),
+    ("aurora", 41.7130, -87.9820),
+    ("argonne", 41.7130, -87.9820),
+    ("anl", 41.7130, -87.9820),
+    ("fugaku", 34.6953, 135.1974),
+    ("kobe", 34.6953, 135.1974),
+    ("riken", 34.6953, 135.1974),
+    ("summit", 35.9319, -84.3107),
+    ("leonardo", 44.4949, 11.3426),
+    ("bologna", 44.4949, 11.3426),
+    ("alps", 46.0037, 8.9511),
+    ("lugano", 46.0037, 8.9511),
+    ("sunway taihulight", 31.4912, 120.3119),
+    ("wuxi", 31.4912, 120.3119),
+    ("tianhe-2", 23.1291, 113.2644),
+    ("tianhe-2a", 23.1291, 113.2644),
+    ("guangzhou", 23.1291, 113.2644),
+    ("colossus", 35.1495, -90.0490),
+    ("memphis", 35.1495, -90.0490),
+    ("xai", 35.1495, -90.0490),
+)
+
+
+def _normalize_hint_text(*parts: Any) -> str:
+    return " ".join(
+        str(part).strip().lower()
+        for part in parts
+        if part not in (None, "")
+    )
+
+
+def _resolve_compute_center_coordinates(
+    record: CollectedData,
+    metadata: Dict[str, Any],
+) -> Dict[str, Any]:
+    latitude = _parse_float(get_record_field(record, "latitude"))
+    longitude = _parse_float(get_record_field(record, "longitude"))
+    if latitude not in (None, 0.0) and longitude not in (None, 0.0):
+        return {
+            "latitude": latitude,
+            "longitude": longitude,
+            "location_precision": "precise",
+            "geography_mode": "source_coordinates",
+            "is_estimated": False,
+            "estimated_reason": None,
+        }
+
+    hint_text = _normalize_hint_text(
+        record.name,
+        get_record_field(record, "city"),
+        get_record_field(record, "country"),
+        metadata.get("site"),
+        metadata.get("organization"),
+        metadata.get("operator"),
+    )
+    for needle, resolved_latitude, resolved_longitude in COMPUTE_CENTER_COORDINATE_HINTS:
+        if needle in hint_text:
+            return {
+                "latitude": resolved_latitude,
+                "longitude": resolved_longitude,
+                "location_precision": "estimated_site",
+                "geography_mode": "site_hint",
+                "is_estimated": True,
+                "estimated_reason": f"Matched known site hint: {needle}",
+            }
+
+    centroid = get_country_centroid(get_record_field(record, "country"))
+    if centroid:
+        return {
+            "latitude": centroid.get("latitude"),
+            "longitude": centroid.get("longitude"),
+            "location_precision": "estimated_country",
+            "geography_mode": "country_centroid",
+            "is_estimated": True,
+            "estimated_reason": "Estimated from country centroid",
+        }
+
+    return {
+        "latitude": latitude,
+        "longitude": longitude,
+        "location_precision": "unknown",
+        "geography_mode": "unknown",
+        "is_estimated": True,
+        "estimated_reason": "No resolvable location hints",
+    }
+
+
+def _normalize_capacity_band(capacity_value: Optional[float], capacity_unit: str) -> str:
+    if capacity_value is None:
+        return "unknown"
+
+    unit = str(capacity_unit or "").strip().lower()
+    if unit in {"pflop/s", "pflops", "pflop"}:
+        normalized_tflops = capacity_value * 1000
+    elif unit in {"gflop/s", "gflops", "gflop"}:
+        normalized_tflops = capacity_value / 1000
+    else:
+        normalized_tflops = capacity_value
+
+    if normalized_tflops >= 1_000_000:
+        return "exascale"
+    if normalized_tflops >= 100_000:
+        return "ultra"
+    if normalized_tflops >= 10_000:
+        return "large"
+    if normalized_tflops > 0:
+        return "regional"
+    return "unknown"
+
+
+def convert_compute_centers_to_geojson(records: List[CollectedData]) -> Dict[str, Any]:
+    """Convert compute infrastructure records into a unified GeoJSON layer."""
+    features = []
+
+    for record in records:
+        metadata = record.extra_data or {}
+        coordinate_info = _resolve_compute_center_coordinates(record, metadata)
+        latitude = coordinate_info.get("latitude")
+        longitude = coordinate_info.get("longitude")
+        site_type = (
+            "supercomputer"
+            if record.source == "top500" or record.data_type == "supercomputer"
+            else "gpu_cluster"
+        )
+        if latitude in (None, 0.0) or longitude in (None, 0.0):
+            continue
+
+        if site_type == "supercomputer":
+            capacity_value = _parse_float(get_record_field(record, "rmax"))
+            capacity_unit = "GFlops"
+        else:
+            capacity_value = _parse_float(get_record_field(record, "value"))
+            capacity_unit = str(get_record_field(record, "unit") or "TFlop/s")
+
+        vendor = (
+            metadata.get("manufacturer")
+            or metadata.get("vendor")
+            or metadata.get("gpu_type")
+        )
+        operator = (
+            metadata.get("organization")
+            or metadata.get("operator")
+            or metadata.get("owner")
+        )
+        rank = metadata.get("rank")
+        if rank in (None, "") and site_type == "supercomputer":
+            rank = get_record_field(record, "rank")
+
+        updated_at = to_iso8601_utc(record.reference_date or record.collected_at)
+
+        features.append(
+            {
+                "type": "Feature",
+                "id": record.id,
+                "geometry": {
+                    "type": "Point",
+                    "coordinates": [longitude or 0, latitude or 0],
+                },
+                "properties": {
+                    "id": record.id,
+                    "source_id": record.source_id,
+                    "name": record.name,
+                    "site_type": site_type,
+                    "country": get_record_field(record, "country"),
+                    "city": get_record_field(record, "city"),
+                    "latitude": latitude,
+                    "longitude": longitude,
+                    "operator": operator,
+                    "vendor": vendor,
+                    "capacity_value": capacity_value,
+                    "capacity_unit": capacity_unit,
+                    "capacity_band": _normalize_capacity_band(capacity_value, capacity_unit),
+                    "rank": rank,
+                    "gpu_count": metadata.get("gpu_count"),
+                    "gpu_type": metadata.get("gpu_type"),
+                    "cores": get_record_field(record, "cores"),
+                    "power": get_record_field(record, "power"),
+                    "source": record.source,
+                    "updated_at": updated_at,
+                    "status": "observed",
+                    "location_precision": coordinate_info.get("location_precision"),
+                    "geography_mode": coordinate_info.get("geography_mode"),
+                    "is_estimated": coordinate_info.get("is_estimated", False),
+                    "estimated_reason": coordinate_info.get("estimated_reason"),
+                    "data_type": "compute_center",
+                    "metadata": metadata,
+                },
+            }
+        )
+
+    return {"type": "FeatureCollection", "features": features}
+
+
 def convert_bgp_anomalies_to_geojson(
     records: List[BGPAnomaly],
     geography_hints: Optional[Dict[str, Dict[str, Any]]] = None,
@@ -722,9 +1014,7 @@ def convert_bgp_incidents_to_geojson(
 async def get_cables_geojson(db: AsyncSession = Depends(get_db)):
     """获取海底电缆 GeoJSON 数据 (LineString)"""
     try:
-        stmt = select(CollectedData).where(CollectedData.source == "arcgis_cables")
-        result = await db.execute(stmt)
-        records = dedupe_collected_records(list(result.scalars().all()))
+        records = await _load_current_collected_data(db, "arcgis_cables")
 
         if not records:
             raise HTTPException(
@@ -736,42 +1026,46 @@ async def get_cables_geojson(db: AsyncSession = Depends(get_db)):
     except HTTPException:
         raise
     except Exception as e:
+        logger.exception_event(
+            "Failed to build cables GeoJSON response",
+            event="visualization.cables.load_failed",
+            context={"error": str(e)},
+        )
+        await record_system_log(
+            source="backend",
+            service="api",
+            module=__name__,
+            event="visualization.cables.load_failed",
+            level="error",
+            message="Failed to build cables GeoJSON response",
+            category="visualization",
+            context={"error": str(e)},
+        )
         raise HTTPException(status_code=500, detail=f"Internal error: {str(e)}")
 
 
 @router.get("/geo/landing-points")
 async def get_landing_points_geojson(db: AsyncSession = Depends(get_db)):
     try:
-        landing_stmt = select(CollectedData).where(CollectedData.source == "arcgis_landing_points")
-        landing_result = await db.execute(landing_stmt)
-        records = dedupe_collected_records(list(landing_result.scalars().all()))
-        
-        relation_stmt = select(CollectedData).where(CollectedData.source == "arcgis_cable_landing_relation")
-        relation_result = await db.execute(relation_stmt)
-        relation_records = dedupe_collected_records(list(relation_result.scalars().all()))
-        
-        cable_stmt = select(CollectedData).where(CollectedData.source == "arcgis_cables")
-        cable_result = await db.execute(cable_stmt)
-        cable_records = dedupe_collected_records(list(cable_result.scalars().all()))
-        
-        city_to_cable_ids_map = {}
-        for rel in relation_records:
-            if rel.extra_data:
-                city_id = rel.extra_data.get("city_id")
-                cable_id = rel.extra_data.get("cable_id")
-                if city_id is not None and cable_id is not None:
-                    if city_id not in city_to_cable_ids_map:
-                        city_to_cable_ids_map[city_id] = []
-                    if cable_id not in city_to_cable_ids_map[city_id]:
-                        city_to_cable_ids_map[city_id].append(cable_id)
-        
-        cable_id_to_name_map = {}
-        for cable in cable_records:
-            if cable.extra_data:
-                cable_id = cable.extra_data.get("cable_id")
-                cable_name = cable.name
-                if cable_id and cable_name:
-                    cable_id_to_name_map[cable_id] = cable_name
+        records_by_source = await _load_current_collected_data_by_sources(
+            db,
+            [
+                "arcgis_landing_points",
+                "arcgis_cable_landing_relation",
+                "arcgis_cables",
+            ],
+        )
+        records = records_by_source.get("arcgis_landing_points", [])
+        relation_records = records_by_source.get(
+            "arcgis_cable_landing_relation",
+            [],
+        )
+        cable_records = records_by_source.get("arcgis_cables", [])
+
+        city_to_cable_ids_map, cable_id_to_name_map = _build_landing_point_cable_maps(
+            relation_records,
+            cable_records,
+        )
         
         if not records:
             raise HTTPException(
@@ -783,41 +1077,85 @@ async def get_landing_points_geojson(db: AsyncSession = Depends(get_db)):
     except HTTPException:
         raise
     except Exception as e:
+        logger.exception_event(
+            "Failed to build landing points GeoJSON response",
+            event="visualization.landing_points.load_failed",
+            context={"error": str(e)},
+        )
+        await record_system_log(
+            source="backend",
+            service="api",
+            module=__name__,
+            event="visualization.landing_points.load_failed",
+            level="error",
+            message="Failed to build landing points GeoJSON response",
+            category="visualization",
+            context={"error": str(e)},
+        )
         raise HTTPException(status_code=500, detail=f"Internal error: {str(e)}")
 
 
+@router.get("/terrain/terrarium/{z}/{x}/{y}.png")
+async def get_terrarium_tile(z: int, x: int, y: int):
+    """Proxy Terrarium elevation tiles through the backend to avoid browser CORS issues."""
+    if z < 0 or x < 0 or y < 0:
+        raise HTTPException(status_code=400, detail="Invalid terrain tile coordinates")
+
+    url = TERRAIN_TILE_URL_TEMPLATE.format(z=z, x=x, y=y)
+
+    try:
+        async with httpx.AsyncClient(
+            timeout=20.0,
+            follow_redirects=True,
+        ) as client:
+            upstream = await client.get(url)
+            upstream.raise_for_status()
+    except httpx.HTTPStatusError as exc:
+        raise HTTPException(
+            status_code=exc.response.status_code,
+            detail=f"Terrain tile upstream error: {exc.response.status_code}",
+        ) from exc
+    except httpx.HTTPError as exc:
+        raise HTTPException(
+            status_code=502,
+            detail=f"Terrain tile fetch failed: {exc}",
+        ) from exc
+
+    cache_control = upstream.headers.get("cache-control") or "public, max-age=86400"
+    etag = upstream.headers.get("etag")
+    last_modified = upstream.headers.get("last-modified")
+    headers = {
+        "Cache-Control": cache_control,
+    }
+    if etag:
+        headers["ETag"] = etag
+    if last_modified:
+        headers["Last-Modified"] = last_modified
+
+    return Response(
+        content=upstream.content,
+        media_type=upstream.headers.get("content-type", "image/png"),
+        headers=headers,
+    )
+
+
 @router.get("/geo/all")
 async def get_all_geojson(db: AsyncSession = Depends(get_db)):
-    cables_stmt = select(CollectedData).where(CollectedData.source == "arcgis_cables")
-    cables_result = await db.execute(cables_stmt)
-    cables_records = dedupe_collected_records(list(cables_result.scalars().all()))
-
-    points_stmt = select(CollectedData).where(CollectedData.source == "arcgis_landing_points")
-    points_result = await db.execute(points_stmt)
-    points_records = dedupe_collected_records(list(points_result.scalars().all()))
-    
-    relation_stmt = select(CollectedData).where(CollectedData.source == "arcgis_cable_landing_relation")
-    relation_result = await db.execute(relation_stmt)
-    relation_records = dedupe_collected_records(list(relation_result.scalars().all()))
-    
-    city_to_cable_ids_map = {}
-    for rel in relation_records:
-        if rel.extra_data:
-            city_id = rel.extra_data.get("city_id")
-            cable_id = rel.extra_data.get("cable_id")
-            if city_id is not None and cable_id is not None:
-                if city_id not in city_to_cable_ids_map:
-                    city_to_cable_ids_map[city_id] = []
-                if cable_id not in city_to_cable_ids_map[city_id]:
-                    city_to_cable_ids_map[city_id].append(cable_id)
-    
-    cable_id_to_name_map = {}
-    for cable in cables_records:
-        if cable.extra_data:
-            cable_id = cable.extra_data.get("cable_id")
-            cable_name = cable.name
-            if cable_id and cable_name:
-                cable_id_to_name_map[cable_id] = cable_name
+    records_by_source = await _load_current_collected_data_by_sources(
+        db,
+        [
+            "arcgis_cables",
+            "arcgis_landing_points",
+            "arcgis_cable_landing_relation",
+        ],
+    )
+    cables_records = records_by_source.get("arcgis_cables", [])
+    points_records = records_by_source.get("arcgis_landing_points", [])
+    relation_records = records_by_source.get("arcgis_cable_landing_relation", [])
+    city_to_cable_ids_map, cable_id_to_name_map = _build_landing_point_cable_maps(
+        relation_records,
+        cables_records,
+    )
 
     cables = (
         convert_cable_to_geojson(cables_records)
@@ -850,17 +1188,12 @@ async def get_satellites_geojson(
     db: AsyncSession = Depends(get_db),
 ):
     """获取卫星 TLE GeoJSON 数据"""
-    stmt = (
-        select(CollectedData)
-        .where(CollectedData.source == "celestrak_tle")
-        .where(CollectedData.name != "Unknown")
-        .order_by(CollectedData.id.desc())
+    records = await _load_current_collected_data(
+        db,
+        "celestrak_tle",
+        exclude_unknown_name=True,
+        limit=limit,
     )
-    result = await db.execute(stmt)
-    records = dedupe_satellite_records(list(result.scalars().all()))
-
-    if limit is not None:
-        records = records[:limit]
 
     if not records:
         return {"type": "FeatureCollection", "features": [], "count": 0}
@@ -878,15 +1211,12 @@ async def get_supercomputers_geojson(
     db: AsyncSession = Depends(get_db),
 ):
     """获取 TOP500 超算中心 GeoJSON 数据"""
-    stmt = (
-        select(CollectedData)
-        .where(CollectedData.source == "top500")
-        .where(CollectedData.name != "Unknown")
-        .order_by(CollectedData.id.desc())
+    records = await _load_current_collected_data(
+        db,
+        "top500",
+        exclude_unknown_name=True,
+        limit=limit,
     )
-    result = await db.execute(stmt)
-    records = dedupe_collected_records(list(result.scalars().all()))
-    records = records[:limit]
 
     if not records:
         return {"type": "FeatureCollection", "features": [], "count": 0}
@@ -904,15 +1234,12 @@ async def get_gpu_clusters_geojson(
     db: AsyncSession = Depends(get_db),
 ):
     """获取 GPU 集群 GeoJSON 数据"""
-    stmt = (
-        select(CollectedData)
-        .where(CollectedData.source == "epoch_ai_gpu")
-        .where(CollectedData.name != "Unknown")
-        .order_by(CollectedData.id.desc())
+    records = await _load_current_collected_data(
+        db,
+        "epoch_ai_gpu",
+        exclude_unknown_name=True,
+        limit=limit,
     )
-    result = await db.execute(stmt)
-    records = dedupe_collected_records(list(result.scalars().all()))
-    records = records[:limit]
 
     if not records:
         return {"type": "FeatureCollection", "features": [], "count": 0}
@@ -924,6 +1251,53 @@ async def get_gpu_clusters_geojson(
     }
 
 
+@router.get("/geo/compute-centers")
+async def get_compute_centers_geojson(
+    limit: int = Query(200, ge=1, le=1000),
+    db: AsyncSession = Depends(get_db),
+):
+    """获取统一算力中心 GeoJSON 数据"""
+    records_by_source = await _load_current_collected_data_by_sources(
+        db,
+        ["top500", "epoch_ai_gpu"],
+    )
+    records = _filter_known_records(
+        records_by_source.get("top500", []) + records_by_source.get("epoch_ai_gpu", []),
+    )
+    if limit is not None:
+        records = records[:limit]
+
+    if not records:
+        return {
+            "type": "FeatureCollection",
+            "features": [],
+            "count": 0,
+            "stats": {
+                "total": 0,
+                "supercomputers": 0,
+                "gpu_clusters": 0,
+            },
+        }
+
+    geojson = convert_compute_centers_to_geojson(records)
+    features = geojson.get("features", [])
+    return {
+        **geojson,
+        "count": len(features),
+        "stats": {
+            "total": len(features),
+            "supercomputers": sum(
+                1 for feature in features
+                if feature.get("properties", {}).get("site_type") == "supercomputer"
+            ),
+            "gpu_clusters": sum(
+                1 for feature in features
+                if feature.get("properties", {}).get("site_type") == "gpu_cluster"
+            ),
+        },
+    }
+
+
 @router.get("/geo/bgp-anomalies")
 async def get_bgp_anomalies_geojson(
     severity: Optional[str] = Query(None),
@@ -979,6 +1353,71 @@ async def get_bgp_collectors_geojson(db: AsyncSession = Depends(get_db)):
     return {**geojson, "count": len(geojson.get("features", []))}
 
 
+@router.get("/geo/summary")
+async def get_visualization_geo_summary(db: AsyncSession = Depends(get_db)):
+    """Return lightweight Earth HUD counts without loading layer GeoJSON payloads."""
+    records_by_source = await _load_current_collected_data_by_sources(
+        db,
+        [
+            "arcgis_cables",
+            "arcgis_landing_points",
+            "celestrak_tle",
+            "top500",
+            "epoch_ai_gpu",
+        ],
+    )
+
+    cables = convert_cable_to_geojson(records_by_source.get("arcgis_cables", []))
+    landing_points = convert_landing_point_to_geojson(
+        records_by_source.get("arcgis_landing_points", []),
+    )
+    satellites = convert_satellite_to_geojson(
+        _filter_known_records(records_by_source.get("celestrak_tle", [])),
+    )
+    compute_centers = convert_compute_centers_to_geojson(
+        _filter_known_records(
+            records_by_source.get("top500", [])
+            + records_by_source.get("epoch_ai_gpu", []),
+        ),
+    )
+    compute_features = compute_centers.get("features", [])
+
+    active_incident_result = await db.execute(
+        select(func.count(BGPIncident.id)).where(BGPIncident.status == "active"),
+    )
+    active_anomaly_result = await db.execute(
+        select(func.count(BGPAnomaly.id)).where(BGPAnomaly.status == "active"),
+    )
+    active_incident_count = int(active_incident_result.scalar() or 0)
+    active_anomaly_count = int(active_anomaly_result.scalar() or 0)
+    bgp_collectors = await build_bgp_collector_coverage(
+        db,
+        source_filter=("ris_live_bgp", "bgpstream_bgp"),
+    )
+
+    return {
+        "generated_at": to_iso8601_utc(datetime.now(UTC)),
+        "stats": {
+            "cable_count": len(cables.get("features", [])),
+            "landing_point_count": len(landing_points.get("features", [])),
+            "satellite_count": len(satellites.get("features", [])),
+            "compute_center_count": len(compute_features),
+            "supercomputer_count": sum(
+                1 for feature in compute_features
+                if feature.get("properties", {}).get("site_type") == "supercomputer"
+            ),
+            "gpu_cluster_count": sum(
+                1 for feature in compute_features
+                if feature.get("properties", {}).get("site_type") == "gpu_cluster"
+            ),
+            "bgp_event_count": active_incident_count or active_anomaly_count,
+            "bgp_incident_count": active_incident_count,
+            "bgp_anomaly_count": active_anomaly_count,
+            "bgp_collector_count": len([item for item in bgp_collectors if item.get("collector")]),
+        },
+    }
+
+
 @router.get("/all")
 async def get_all_visualization_data(db: AsyncSession = Depends(get_db)):
     """获取所有可视化数据的统一端点
@@ -990,37 +1429,27 @@ async def get_all_visualization_data(db: AsyncSession = Depends(get_db)):
     - supercomputers: TOP500 超算
     - gpu_clusters: GPU 集群
     """
-    cables_stmt = select(CollectedData).where(CollectedData.source == "arcgis_cables")
-    cables_result = await db.execute(cables_stmt)
-    cables_records = dedupe_collected_records(list(cables_result.scalars().all()))
-
-    points_stmt = select(CollectedData).where(CollectedData.source == "arcgis_landing_points")
-    points_result = await db.execute(points_stmt)
-    points_records = dedupe_collected_records(list(points_result.scalars().all()))
-
-    satellites_stmt = (
-        select(CollectedData)
-        .where(CollectedData.source == "celestrak_tle")
-        .where(CollectedData.name != "Unknown")
+    records_by_source = await _load_current_collected_data_by_sources(
+        db,
+        [
+            "arcgis_cables",
+            "arcgis_landing_points",
+            "celestrak_tle",
+            "top500",
+            "epoch_ai_gpu",
+        ],
     )
-    satellites_result = await db.execute(satellites_stmt)
-    satellites_records = dedupe_satellite_records(list(satellites_result.scalars().all()))
-
-    supercomputers_stmt = (
-        select(CollectedData)
-        .where(CollectedData.source == "top500")
-        .where(CollectedData.name != "Unknown")
+    cables_records = records_by_source.get("arcgis_cables", [])
+    points_records = records_by_source.get("arcgis_landing_points", [])
+    satellites_records = _filter_known_records(
+        records_by_source.get("celestrak_tle", []),
     )
-    supercomputers_result = await db.execute(supercomputers_stmt)
-    supercomputers_records = dedupe_collected_records(list(supercomputers_result.scalars().all()))
-
-    gpu_stmt = (
-        select(CollectedData)
-        .where(CollectedData.source == "epoch_ai_gpu")
-        .where(CollectedData.name != "Unknown")
+    supercomputers_records = _filter_known_records(
+        records_by_source.get("top500", []),
+    )
+    gpu_records = _filter_known_records(
+        records_by_source.get("epoch_ai_gpu", []),
     )
-    gpu_result = await db.execute(gpu_stmt)
-    gpu_records = dedupe_collected_records(list(gpu_result.scalars().all()))
 
     cables = (
         convert_cable_to_geojson(cables_records)
@@ -1084,13 +1513,8 @@ async def get_cable_graph(db: AsyncSession) -> CableGraph:
     global _cable_graph
 
     if _cable_graph is None:
-        cables_stmt = select(CollectedData).where(CollectedData.source == "arcgis_cables")
-        cables_result = await db.execute(cables_stmt)
-        cables_records = list(cables_result.scalars().all())
-
-        points_stmt = select(CollectedData).where(CollectedData.source == "arcgis_landing_points")
-        points_result = await db.execute(points_stmt)
-        points_records = list(points_result.scalars().all())
+        cables_records = await _load_current_collected_data(db, "arcgis_cables")
+        points_records = await _load_current_collected_data(db, "arcgis_landing_points")
 
         cables_data = convert_cable_to_geojson(cables_records)
         points_data = convert_landing_point_to_geojson(points_records)

backend/app/api/v1/websocket.py

@@ -2,7 +2,6 @@
 
 import asyncio
 import json
-import logging
 from datetime import UTC, datetime
 from typing import Optional
 
@@ -10,10 +9,11 @@ from fastapi import APIRouter, WebSocket, WebSocketDisconnect, Query
 from jose import jwt, JWTError
 
 from app.core.config import settings
+from app.core.logging import get_logger
 from app.core.time import to_iso8601_utc
 from app.core.websocket.manager import manager
 
-logger = logging.getLogger(__name__)
+logger = get_logger(__name__, service="api")
 router = APIRouter()
 
 
@@ -22,11 +22,18 @@ async def authenticate_token(token: str) -> Optional[dict]:
     try:
         payload = jwt.decode(token, settings.SECRET_KEY, algorithms=[settings.ALGORITHM])
         if payload.get("type") != "access":
-            logger.warning(f"WebSocket auth failed: wrong token type")
+            logger.warning_event(
+                "WebSocket auth failed: wrong token type",
+                event="auth.websocket.invalid_token_type",
+            )
             return None
         return payload
     except JWTError as e:
-        logger.warning(f"WebSocket auth failed: {e}")
+        logger.warning_event(
+            "WebSocket auth failed",
+            event="auth.websocket.decode_failed",
+            context={"error": str(e)},
+        )
         return None
 
 
@@ -36,10 +43,17 @@ async def websocket_endpoint(
     token: str = Query(...),
 ):
     """WebSocket endpoint for real-time data"""
-    logger.info(f"WebSocket connection attempt with token: {token[:20]}...")
+    logger.info_event(
+        "WebSocket connection attempt",
+        event="auth.websocket.connection_attempt",
+        context={"token_preview": f"{token[:8]}..."},
+    )
     payload = await authenticate_token(token)
     if payload is None:
-        logger.warning("WebSocket authentication failed, closing connection")
+        logger.warning_event(
+            "WebSocket authentication failed, closing connection",
+            event="auth.websocket.connection_rejected",
+        )
         await websocket.close(code=4001)
         return
 

backend/app/core/cache.py

@@ -1,15 +1,15 @@
 """Redis caching service"""
 
 import json
-import logging
 from datetime import timedelta
 from typing import Optional, Any
 
 import redis
 
 from app.core.config import settings
+from app.core.logging import get_logger
 
-logger = logging.getLogger(__name__)
+logger = get_logger(__name__)
 
 
 # Lazy Redis client initialization
@@ -47,7 +47,7 @@ class CacheService:
                 return json.loads(value)
             return None
         except Exception as e:
-            logger.warning(f"Cache get error: {e}")
+            logger.warning_event("Cache get error", event="cache.get.failed", context={"error": str(e)})
             return None
 
     def set(
@@ -61,7 +61,7 @@ class CacheService:
             serialized = json.dumps(value, default=str)
             return self.client.setex(key, expire_seconds, serialized)
         except Exception as e:
-            logger.warning(f"Cache set error: {e}")
+            logger.warning_event("Cache set error", event="cache.set.failed", context={"error": str(e)})
             return False
 
     def delete(self, key: str) -> bool:
@@ -69,7 +69,7 @@ class CacheService:
         try:
             return self.client.delete(key) > 0
         except Exception as e:
-            logger.warning(f"Cache delete error: {e}")
+            logger.warning_event("Cache delete error", event="cache.delete.failed", context={"error": str(e)})
             return False
 
     def delete_pattern(self, pattern: str) -> int:
@@ -80,7 +80,7 @@ class CacheService:
                 return self.client.delete(*keys)
             return 0
         except Exception as e:
-            logger.warning(f"Cache delete_pattern error: {e}")
+            logger.warning_event("Cache delete_pattern error", event="cache.delete_pattern.failed", context={"error": str(e)})
             return 0
 
     def get_or_set(

backend/app/core/data_sources.py

@@ -11,6 +11,7 @@ COLLECTOR_URL_KEYS = {
     "fao_landing_points": "fao.landing_point_url",
     "telegeography_cables": "telegeography.cable_url",
     "telegeography_landing": "telegeography.landing_point_url",
+    "telegeography_systems": "telegeography.cable_url",
     "huggingface_models": "huggingface.models_url",
     "huggingface_datasets": "huggingface.datasets_url",
     "huggingface_spaces": "huggingface.spaces_url",
@@ -23,11 +24,13 @@ COLLECTOR_URL_KEYS = {
     "top500": "top500.url",
     "epoch_ai_gpu": "epoch_ai.gpu_clusters_url",
     "spacetrack_tle": "spacetrack.tle_query_url",
+    "celestrak_tle": "celestrak.base_url",
     "ris_live_bgp": "ris_live.url",
     "bgpstream_bgp": "bgpstream.url",
     "iptoasn_prefix_geo": "iptoasn.combined_url",
     "opengeofeed_prefix_geo": "opengeofeed.public_csv_url",
     "nro_delegated_prefix_geo": "nro.delegated_stats_url",
+    "news_live_streams": "news_live_streams.channels_url",
 }
 
 
@@ -41,18 +44,22 @@ class DataSourcesConfig:
             with open(config_path, "r") as f:
                 self._yaml_config = yaml.safe_load(f) or {}
 
-    def get_yaml_url(self, collector_name: str) -> str:
-        key = COLLECTOR_URL_KEYS.get(collector_name, "")
+    def get_yaml_value(self, key: str):
         if not key:
-            return ""
+            return None
 
         parts = key.split(".")
         value = self._yaml_config
         for part in parts:
             if isinstance(value, dict):
-                value = value.get(part, "")
+                value = value.get(part)
             else:
-                return ""
+                return None
+        return value
+
+    def get_yaml_url(self, collector_name: str) -> str:
+        key = COLLECTOR_URL_KEYS.get(collector_name, "")
+        value = self.get_yaml_value(key)
         return value if isinstance(value, str) else ""
 
     async def get_url(self, collector_name: str, db) -> str:

backend/app/core/data_sources.yaml

@@ -2,53 +2,95 @@
 # All external data source URLs should be configured here
 
 arcgis:
+  # ArcGIS 海缆 GeoJSON 查询接口
   cable_url: "https://services.arcgis.com/6DIQcwlPy8knb6sg/ArcGIS/rest/services/SubmarineCables/FeatureServer/2/query"
+  # ArcGIS 登陆点 GeoJSON 查询接口
   landing_point_url: "https://services.arcgis.com/6DIQcwlPy8knb6sg/ArcGIS/rest/services/SubmarineCables/FeatureServer/1/query"
+  # ArcGIS 海缆与登陆点关联关系查询接口
   cable_landing_relation_url: "https://services.arcgis.com/6DIQcwlPy8knb6sg/ArcGIS/rest/services/SubmarineCables/FeatureServer/3/query"
 
 fao:
+  # FAO 登陆点 CSV 下载地址
   landing_point_url: "https://data.apps.fao.org/catalog/dataset/1b75ff21-92f2-4b96-9b7b-98e8aa65ad5d/resource/b6071077-d1d4-4e97-aa00-42e902847c87/download/landing-point-geo.csv"
 
 telegeography:
+  # TeleGeography 海缆/系统主数据源，当前使用 GitHub 镜像 JSON
   cable_url: "https://raw.githubusercontent.com/lintaojlu/submarine_cable_information/main/cable.json"
+  # TeleGeography 登陆点主数据源，当前使用 GitHub 镜像 JSON
   landing_point_url: "https://raw.githubusercontent.com/lintaojlu/submarine_cable_information/main/landing_point.json"
+  # TeleGeography 历史 API 存档，用于 cable collector 的 fallback
+  archived_cable_url: "https://web.archive.org/web/2024/https://www.submarinecablemap.com/api/v3/cable"
+  # TeleGeography 官网页面，用于 cable collector 的最终 HTML 抓取 fallback
+  live_map_url: "https://www.submarinecablemap.com"
 
 huggingface:
+  # Hugging Face 模型目录 API
   models_url: "https://huggingface.co/api/models"
+  # Hugging Face 数据集目录 API
   datasets_url: "https://huggingface.co/api/datasets"
+  # Hugging Face Spaces 目录 API
   spaces_url: "https://huggingface.co/api/spaces"
 
 cloudflare:
+  # Cloudflare Radar 设备类型摘要接口
   radar_device_url: "https://api.cloudflare.com/client/v4/radar/http/summary/device_type"
+  # Cloudflare Radar 请求量时间序列接口
   radar_traffic_url: "https://api.cloudflare.com/client/v4/radar/http/timeseries/requests"
+  # Cloudflare Radar 热点地理位置接口
   radar_top_locations_url: "https://api.cloudflare.com/client/v4/radar/http/top/locations"
 
 peeringdb:
+  # PeeringDB IXP API
   ixp_url: "https://www.peeringdb.com/api/ix"
+  # PeeringDB Network API
   network_url: "https://www.peeringdb.com/api/net"
+  # PeeringDB Facility API
   facility_url: "https://www.peeringdb.com/api/fac"
 
 top500:
+  # TOP500 榜单页面，用于主表抓取
   url: "https://top500.org/lists/top500/list/2025/11/"
+  # TOP500 站点根地址，用于拼详情页链接
+  base_url: "https://top500.org"
 
 epoch_ai:
+  # Epoch AI GPU Cluster 页面
   gpu_clusters_url: "https://epoch.ai/data/gpu-clusters"
 
 spacetrack:
+  # Space-Track 站点根地址，用于首页访问和登录地址推导
   base_url: "https://www.space-track.org"
+  # Space-Track TLE 主查询接口
   tle_query_url: "https://www.space-track.org/basicspacedata/query/class/gp/orderby/EPOCH%20desc/limit/1000/format/json"
 
+celestrak:
+  # CelesTrak TLE 基础接口，collector 会在其后拼接 GROUP / FORMAT 参数
+  base_url: "https://celestrak.org/NORAD/elements/gp.php"
+
 ris_live:
+  # RIPE RIS Live 流式订阅地址
   url: "https://ris-live.ripe.net/v1/stream/?format=json&client=planet-ris-live"
 
 bgpstream:
+  # CAIDA BGPStream Broker API
   url: "https://broker.bgpstream.caida.org/v2"
 
 iptoasn:
+  # IPtoASN prefix geography 合并数据下载地址
   combined_url: "https://iptoasn.com/data/ip2asn-combined.tsv.gz"
 
 opengeofeed:
+  # OpenGeoFeed 公共 geofeed CSV
   public_csv_url: "https://opengeofeed.org/feed/public.csv"
 
 nro:
+  # NRO delegated stats 下载地址
   delegated_stats_url: "https://ftp.ripe.net/pub/stats/ripencc/nro-stats/latest/nro-delegated-stats"
+
+news_live_streams:
+  # IPTV-org 频道元数据 JSON
+  channels_url: "https://iptv-org.github.io/api/channels.json"
+  # IPTV-org 频道播放流 JSON
+  streams_url: "https://iptv-org.github.io/api/streams.json"
+  # IPTV-org 台标 JSON
+  logos_url: "https://iptv-org.github.io/api/logos.json"

backend/app/core/datasource_defaults.py

@@ -155,6 +155,13 @@ DEFAULT_DATASOURCES = {
         "priority": "P1",
         "frequency_minutes": 1440,
     },
+    "news_live_streams": {
+        "id": 26,
+        "name": "News Live Streams",
+        "module": "L4",
+        "priority": "P2",
+        "frequency_minutes": 720,
+    },
 }
 
 ID_TO_COLLECTOR = {info["id"]: name for name, info in DEFAULT_DATASOURCES.items()}

backend/app/core/logging.py

@@ -0,0 +1,161 @@
+from __future__ import annotations
+
+import json
+import logging
+import os
+import re
+
+from collections.abc import Mapping, Sequence
+from typing import Any
+
+from app.core.request_context import get_request_id
+
+DEFAULT_SERVICE = "backend"
+DEFAULT_EVENT = "app.log"
+DEFAULT_LOG_LEVEL = os.getenv("PLANET_LOG_LEVEL", "INFO").upper()
+REDACTED = "[REDACTED]"
+SENSITIVE_FIELD_NAMES = {
+    "access_token",
+    "api_key",
+    "authorization",
+    "cookie",
+    "password",
+    "refresh_token",
+    "secret",
+    "token",
+}
+SENSITIVE_TEXT_PATTERNS = (
+    re.compile(r"(?i)(authorization\s*[:=]\s*)(.+)"),
+    re.compile(r"(?i)(bearer\s+)([A-Za-z0-9._\-]+)"),
+    re.compile(r"(?i)(token\s*[:=]\s*)(.+)"),
+    re.compile(r"(?i)(password\s*[:=]\s*)(.+)"),
+    re.compile(r"(?i)(cookie\s*[:=]\s*)(.+)"),
+)
+
+
+def sanitize_log_value(value: Any) -> Any:
+    if isinstance(value, Mapping):
+        return {
+            str(key): (REDACTED if str(key).lower() in SENSITIVE_FIELD_NAMES else sanitize_log_value(item))
+            for key, item in value.items()
+        }
+    if isinstance(value, Sequence) and not isinstance(value, (str, bytes, bytearray)):
+        return [sanitize_log_value(item) for item in value]
+    if isinstance(value, str):
+        sanitized = value
+        for pattern in SENSITIVE_TEXT_PATTERNS:
+            sanitized = pattern.sub(lambda match: f"{match.group(1)}{REDACTED}", sanitized)
+        return sanitized
+    return value
+
+
+def _normalize_context(context: Any) -> dict[str, Any]:
+    if context is None:
+        return {}
+    if isinstance(context, Mapping):
+        sanitized = sanitize_log_value(context)
+        return {str(key): value for key, value in sanitized.items()}
+    return {"value": sanitize_log_value(context)}
+
+
+class PlanetContextFilter(logging.Filter):
+    def filter(self, record: logging.LogRecord) -> bool:
+        record.request_id = getattr(record, "request_id", None) or get_request_id() or "-"
+        record.service = getattr(record, "service", None) or DEFAULT_SERVICE
+        record.event = getattr(record, "event", None) or DEFAULT_EVENT
+        record.context = _normalize_context(getattr(record, "context", None))
+        record.message = sanitize_log_value(record.getMessage())
+        return True
+
+
+class PlanetFormatter(logging.Formatter):
+    def format(self, record: logging.LogRecord) -> str:
+        timestamp = self.formatTime(record, self.datefmt)
+        level = record.levelname
+        service = getattr(record, "service", DEFAULT_SERVICE)
+        module_name = record.name
+        event = getattr(record, "event", DEFAULT_EVENT)
+        request_id = getattr(record, "request_id", "-")
+        message = sanitize_log_value(record.getMessage())
+        context = _normalize_context(getattr(record, "context", None))
+        context_suffix = ""
+        if context:
+            context_suffix = f" context={json.dumps(context, ensure_ascii=False, sort_keys=True)}"
+        rendered = (
+            f"{timestamp} {level} service={service} module={module_name} "
+            f"event={event} request_id={request_id} message={message}{context_suffix}"
+        )
+        if record.exc_info:
+            rendered = f"{rendered}\n{self.formatException(record.exc_info)}"
+        return rendered
+
+
+class PlanetLoggerAdapter(logging.LoggerAdapter):
+    def process(self, msg: Any, kwargs: dict[str, Any]) -> tuple[Any, dict[str, Any]]:
+        extra = dict(self.extra)
+        extra.update(kwargs.get("extra", {}))
+        if "context" in extra:
+            extra["context"] = _normalize_context(extra.get("context"))
+        kwargs["extra"] = extra
+        return sanitize_log_value(msg), kwargs
+
+    def log_event(
+        self,
+        level: int,
+        message: str,
+        *,
+        event: str,
+        context: Mapping[str, Any] | None = None,
+        **extra: Any,
+    ) -> None:
+        self.log(level, message, extra={"event": event, "context": context or {}, **extra})
+
+    def debug_event(self, message: str, *, event: str, context: Mapping[str, Any] | None = None, **extra: Any) -> None:
+        self.log_event(logging.DEBUG, message, event=event, context=context, **extra)
+
+    def info_event(self, message: str, *, event: str, context: Mapping[str, Any] | None = None, **extra: Any) -> None:
+        self.log_event(logging.INFO, message, event=event, context=context, **extra)
+
+    def warning_event(self, message: str, *, event: str, context: Mapping[str, Any] | None = None, **extra: Any) -> None:
+        self.log_event(logging.WARNING, message, event=event, context=context, **extra)
+
+    def error_event(self, message: str, *, event: str, context: Mapping[str, Any] | None = None, **extra: Any) -> None:
+        self.log_event(logging.ERROR, message, event=event, context=context, **extra)
+
+    def exception_event(
+        self,
+        message: str,
+        *,
+        event: str,
+        context: Mapping[str, Any] | None = None,
+        **extra: Any,
+    ) -> None:
+        self.error(message, exc_info=True, extra={"event": event, "context": context or {}, **extra})
+
+
+def get_logger(name: str, *, service: str = DEFAULT_SERVICE) -> PlanetLoggerAdapter:
+    return PlanetLoggerAdapter(logging.getLogger(name), {"service": service})
+
+
+def configure_logging(level: str | None = None) -> None:
+    root_logger = logging.getLogger()
+    if getattr(configure_logging, "_configured", False):
+        if level:
+            root_logger.setLevel(level.upper())
+        return
+
+    handler = logging.StreamHandler()
+    handler.setFormatter(PlanetFormatter(datefmt="%Y-%m-%d %H:%M:%S"))
+    handler.addFilter(PlanetContextFilter())
+
+    root_logger.handlers.clear()
+    root_logger.addHandler(handler)
+    root_logger.setLevel((level or DEFAULT_LOG_LEVEL).upper())
+
+    for logger_name in ("uvicorn", "uvicorn.error", "uvicorn.access"):
+        target_logger = logging.getLogger(logger_name)
+        target_logger.handlers.clear()
+        target_logger.propagate = True
+
+    logging.captureWarnings(True)
+    configure_logging._configured = True

backend/app/core/request_context.py

@@ -0,0 +1,14 @@
+from __future__ import annotations
+
+from contextvars import ContextVar
+
+
+request_id_context: ContextVar[str | None] = ContextVar("request_id", default=None)
+
+
+def set_request_id(request_id: str | None) -> None:
+    request_id_context.set(request_id)
+
+
+def get_request_id() -> str | None:
+    return request_id_context.get()

backend/app/db/session.py

@@ -5,10 +5,22 @@ from sqlalchemy.ext.asyncio import AsyncSession, create_async_engine, async_sess
 from sqlalchemy.orm import declarative_base
 
 from app.core.config import settings
+from app.core.logging import get_logger
+
+logger = get_logger(__name__)
+
+DB_POOL_CONFIG = {
+    "pool_pre_ping": True,
+    "pool_recycle": 1800,
+    "pool_size": 10,
+    "max_overflow": 20,
+    "pool_timeout": 30,
+}
 
 engine = create_async_engine(
     settings.DATABASE_URL,
     echo=settings.DEBUG if hasattr(settings, "DEBUG") else False,
+    **DB_POOL_CONFIG,
 )
 
 async_session_factory = async_sessionmaker(engine, class_=AsyncSession, expire_on_commit=False)
@@ -95,6 +107,21 @@ async def init_db():
     import app.models.bgp_observation  # noqa: F401
     import app.models.collected_data  # noqa: F401
     import app.models.system_setting  # noqa: F401
+    import app.models.playground_session  # noqa: F401
+    import app.models.playground_message  # noqa: F401
+    import app.models.system_log  # noqa: F401
+
+    logger.warning_event(
+        "Database pool settings active",
+        event="database.pool.initialized",
+        context={
+            "pool_pre_ping": DB_POOL_CONFIG["pool_pre_ping"],
+            "pool_recycle": DB_POOL_CONFIG["pool_recycle"],
+            "pool_size": DB_POOL_CONFIG["pool_size"],
+            "max_overflow": DB_POOL_CONFIG["max_overflow"],
+            "pool_timeout": DB_POOL_CONFIG["pool_timeout"],
+        },
+    )
 
     async with engine.begin() as conn:
         await conn.run_sync(Base.metadata.create_all)

backend/app/main.py

@@ -1,4 +1,5 @@
 from contextlib import asynccontextmanager
+from uuid import uuid4
 
 from fastapi import FastAPI
 from fastapi.middleware.cors import CORSMiddleware
@@ -7,6 +8,8 @@ from starlette.middleware.base import BaseHTTPMiddleware
 from app.api.main import api_router
 from app.api.v1 import websocket
 from app.core.config import settings
+from app.core.logging import configure_logging
+from app.core.request_context import set_request_id
 from app.core.websocket.broadcaster import broadcaster
 from app.db.session import init_db
 from app.services.scheduler import (
@@ -17,6 +20,9 @@ from app.services.scheduler import (
 )
 
 
+configure_logging()
+
+
 class WebSocketCORSMiddleware(BaseHTTPMiddleware):
     async def dispatch(self, request, call_next):
         if request.url.path.startswith("/ws") and request.method == "GET":
@@ -28,6 +34,18 @@ class WebSocketCORSMiddleware(BaseHTTPMiddleware):
         return await call_next(request)
 
 
+class RequestContextMiddleware(BaseHTTPMiddleware):
+    async def dispatch(self, request, call_next):
+        request_id = request.headers.get("X-Request-ID") or uuid4().hex
+        set_request_id(request_id)
+        try:
+            response = await call_next(request)
+        finally:
+            set_request_id(None)
+        response.headers["X-Request-ID"] = request_id
+        return response
+
+
 @asynccontextmanager
 async def lifespan(app: FastAPI):
     await init_db()
@@ -58,6 +76,7 @@ app.add_middleware(
     allow_headers=["*"],
 )
 
+app.add_middleware(RequestContextMiddleware)
 app.add_middleware(WebSocketCORSMiddleware)
 
 app.include_router(api_router, prefix="/api/v1")

backend/app/models/__init__.py

@@ -9,6 +9,9 @@ from app.models.bgp_anomaly import BGPAnomaly
 from app.models.bgp_incident import BGPIncident
 from app.models.bgp_observation import BGPObservation
 from app.models.system_setting import SystemSetting
+from app.models.playground_session import PlaygroundSession
+from app.models.playground_message import PlaygroundMessage
+from app.models.system_log import SystemLog, AuditLog
 
 __all__ = [
     "User",
@@ -24,4 +27,6 @@ __all__ = [
     "BGPAnomaly",
     "BGPIncident",
     "BGPObservation",
+    "SystemLog",
+    "AuditLog",
 ]

backend/app/models/playground_message.py

@@ -0,0 +1,40 @@
+from sqlalchemy import JSON, Boolean, Column, DateTime, ForeignKey, Integer, String, Text
+from sqlalchemy.sql import func
+
+from app.db.session import Base
+
+
+class PlaygroundMessage(Base):
+    __tablename__ = "playground_messages"
+
+    id = Column(Integer, primary_key=True, autoincrement=True)
+    public_id = Column(String(64), unique=True, index=True, nullable=False)
+    session_id = Column(Integer, ForeignKey("playground_sessions.id", ondelete="CASCADE"), nullable=False, index=True)
+    user_id = Column(Integer, ForeignKey("users.id", ondelete="CASCADE"), nullable=False, index=True)
+    parent_message_id = Column(Integer, ForeignKey("playground_messages.id", ondelete="SET NULL"), nullable=True)
+    role = Column(String(20), nullable=False)
+    kind = Column(String(20), nullable=False, default="message")
+    status = Column(String(20), nullable=False, default="done")
+    title = Column(String(255), nullable=True)
+    content = Column(Text, nullable=False, default="")
+    thinking_content = Column(Text, nullable=False, default="")
+    meta = Column(JSON, nullable=False, default=list)
+    provider = Column(String(100), nullable=True)
+    model = Column(String(200), nullable=True)
+    request_id = Column(String(100), nullable=True)
+    raw_response = Column(JSON, nullable=False, default=dict)
+    content_blocks = Column(JSON, nullable=False, default=list)
+    text_blocks = Column(JSON, nullable=False, default=list)
+    thinking_blocks = Column(JSON, nullable=False, default=list)
+    sort_order = Column(Integer, nullable=False, default=0, index=True)
+    is_visible = Column(Boolean, nullable=False, default=True)
+    created_at = Column(DateTime(timezone=True), server_default=func.now(), nullable=False)
+    updated_at = Column(
+        DateTime(timezone=True),
+        server_default=func.now(),
+        onupdate=func.now(),
+        nullable=False,
+    )
+
+    def __repr__(self):
+        return f"<PlaygroundMessage public_id={self.public_id} role={self.role} status={self.status}>"

backend/app/models/playground_session.py

@@ -0,0 +1,27 @@
+from sqlalchemy import JSON, Column, DateTime, ForeignKey, Integer, String, UniqueConstraint
+from sqlalchemy.sql import func
+
+from app.db.session import Base
+
+
+class PlaygroundSession(Base):
+    __tablename__ = "playground_sessions"
+    __table_args__ = (
+        UniqueConstraint("user_id", "session_key", name="uq_playground_sessions_user_session_key"),
+    )
+
+    id = Column(Integer, primary_key=True, autoincrement=True)
+    user_id = Column(Integer, ForeignKey("users.id", ondelete="CASCADE"), nullable=False, index=True)
+    session_key = Column(String(100), nullable=False, default="default")
+    title = Column(String(200), nullable=False, default="Playground 会话")
+    state = Column(JSON, nullable=False, default={})
+    created_at = Column(DateTime(timezone=True), server_default=func.now(), nullable=False)
+    updated_at = Column(
+        DateTime(timezone=True),
+        server_default=func.now(),
+        onupdate=func.now(),
+        nullable=False,
+    )
+
+    def __repr__(self):
+        return f"<PlaygroundSession user_id={self.user_id} session_key={self.session_key}>"

backend/app/models/system_log.py

@@ -0,0 +1,40 @@
+from sqlalchemy import JSON, Column, DateTime, Integer, String, Text
+from sqlalchemy.sql import func
+
+from app.db.session import Base
+
+
+class SystemLog(Base):
+    __tablename__ = "system_logs"
+
+    id = Column(Integer, primary_key=True, autoincrement=True)
+    occurred_at = Column(DateTime(timezone=True), server_default=func.now(), index=True)
+    source = Column(String(50), nullable=False, index=True)
+    service = Column(String(50), nullable=True)
+    module = Column(String(120), nullable=True)
+    event = Column(String(160), nullable=True, index=True)
+    level = Column(String(20), nullable=False, index=True)
+    message = Column(Text, nullable=False)
+    request_id = Column(String(64), nullable=True, index=True)
+    trace_id = Column(String(64), nullable=True)
+    user_id = Column(Integer, nullable=True, index=True)
+    category = Column(String(80), nullable=True, index=True)
+    context = Column(JSON, nullable=False, default=dict)
+    created_at = Column(DateTime(timezone=True), server_default=func.now())
+
+
+class AuditLog(Base):
+    __tablename__ = "audit_logs"
+
+    id = Column(Integer, primary_key=True, autoincrement=True)
+    occurred_at = Column(DateTime(timezone=True), server_default=func.now(), index=True)
+    actor_id = Column(Integer, nullable=True, index=True)
+    actor_name = Column(String(255), nullable=True)
+    action = Column(String(120), nullable=False, index=True)
+    target_type = Column(String(80), nullable=True)
+    target_id = Column(String(120), nullable=True)
+    result = Column(String(40), nullable=True, index=True)
+    request_id = Column(String(64), nullable=True, index=True)
+    ip = Column(String(64), nullable=True)
+    details = Column(JSON, nullable=False, default=dict)
+    created_at = Column(DateTime(timezone=True), server_default=func.now())

backend/app/schemas/ai.py

@@ -3,6 +3,14 @@ from typing import Any
 from pydantic import BaseModel, Field
 
 
+class AIContentBlock(BaseModel):
+    type: str
+    text: str | None = None
+    thinking: str | None = None
+    signature: str | None = None
+    metadata: dict[str, Any] = Field(default_factory=dict)
+
+
 class SituationalAnalysisRequest(BaseModel):
     title: str = Field(..., min_length=1, max_length=200)
     objective: str = Field(..., min_length=1, max_length=1000)
@@ -10,18 +18,158 @@ class SituationalAnalysisRequest(BaseModel):
     observations: list[str] = Field(default_factory=list)
     constraints: list[str] = Field(default_factory=list)
     preferred_model: str | None = Field(default=None, max_length=200)
+    thinking: dict[str, Any] | None = None
+
+
+class BGPBriefRequest(BaseModel):
+    incident_limit: int = Field(default=5, ge=1, le=10)
+    anomaly_limit: int = Field(default=6, ge=1, le=12)
+    collector_limit: int = Field(default=5, ge=1, le=10)
+    preferred_model: str | None = Field(default=None, max_length=200)
+    thinking: dict[str, Any] | None = None
+
+
+class AlertBriefRequest(BaseModel):
+    alert_limit: int = Field(default=8, ge=1, le=20)
+    preferred_model: str | None = Field(default=None, max_length=200)
+    thinking: dict[str, Any] | None = None
+
+
+class SituationalAlertBriefRequest(BaseModel):
+    preferred_model: str | None = Field(default=None, max_length=200)
+    thinking: dict[str, Any] | None = None
 
 
 class SituationalAnalysisResponse(BaseModel):
     provider: str
     model: str
     content: str
+    content_blocks: list[AIContentBlock] = Field(default_factory=list)
+    text_blocks: list[str] = Field(default_factory=list)
+    thinking_blocks: list[str] = Field(default_factory=list)
     raw_response: dict[str, Any] = Field(default_factory=dict)
 
 
+class BGPBriefRecordSummary(BaseModel):
+    id: str
+    title: str
+    provider: str
+    model: str
+    request_id: str | None = None
+    generated_at: str
+
+
+class BGPBriefRecordResponse(BGPBriefRecordSummary):
+    content_markdown: str
+    facts: list[str] = Field(default_factory=list)
+    context: dict[str, Any] = Field(default_factory=dict)
+
+
+class AlertBriefResponse(SituationalAnalysisResponse):
+    title: str
+    objective: str
+    facts: list[str] = Field(default_factory=list)
+    context: dict[str, Any] = Field(default_factory=dict)
+
+
+class SituationalAlertBriefResponse(SituationalAnalysisResponse):
+    title: str
+    objective: str
+    facts: list[str] = Field(default_factory=list)
+    context: dict[str, Any] = Field(default_factory=dict)
+
+
 class AIProviderStatusResponse(BaseModel):
     provider: str
+    api: str | None = None
     enabled: bool
     configured: bool
     model: str | None = None
     base_url: str | None = None
+
+
+class PlaygroundSessionState(BaseModel):
+    messages: list[dict[str, Any]] = Field(default_factory=list)
+    selectedPresetKey: str = Field(default="bgp-brief", max_length=100)
+    title: str = Field(default="", max_length=200)
+    objective: str = Field(default="", max_length=1000)
+    constraints: str = Field(default="")
+    inputValue: str = Field(default="")
+    analysis: dict[str, Any] | None = None
+    latestAnalysisMessageId: str | None = Field(default=None, max_length=200)
+    analysisMeta: dict[str, Any] = Field(default_factory=dict)
+    helpExpanded: bool = True
+
+
+class PlaygroundSessionUpsertRequest(BaseModel):
+    session_key: str = Field(default="default", min_length=1, max_length=100)
+    title: str | None = Field(default=None, max_length=200)
+    state: PlaygroundSessionState
+
+
+class PlaygroundMessageRecord(BaseModel):
+    id: str
+    role: str
+    kind: str = "message"
+    status: str = "done"
+    title: str | None = None
+    content: str = ""
+    thinking_content: str = ""
+    meta: list[str] = Field(default_factory=list)
+    markdown: bool = True
+    provider: str | None = None
+    model: str | None = None
+    request_id: str | None = None
+    raw_response: dict[str, Any] = Field(default_factory=dict)
+    content_blocks: list[dict[str, Any]] = Field(default_factory=list)
+    text_blocks: list[str] = Field(default_factory=list)
+    thinking_blocks: list[str] = Field(default_factory=list)
+    parent_message_id: str | None = None
+    created_at: str
+    updated_at: str
+
+
+class PlaygroundSessionResponse(BaseModel):
+    id: str
+    session_key: str
+    title: str
+    state: PlaygroundSessionState
+    created_at: str
+    updated_at: str
+
+
+class PlaygroundThreadResponse(BaseModel):
+    session: PlaygroundSessionResponse
+    messages: list[PlaygroundMessageRecord] = Field(default_factory=list)
+
+
+class PlaygroundMessageCreateRequest(BaseModel):
+    session_key: str = Field(default="default", min_length=1, max_length=100)
+    title: str = Field(..., min_length=1, max_length=200)
+    objective: str = Field(..., min_length=1, max_length=1000)
+    constraints: str = Field(default="")
+    input: str = Field(..., min_length=1)
+    selected_preset_key: str = Field(default="bgp-brief", max_length=100)
+    help_expanded: bool = True
+
+
+class PlaygroundMessageActionResponse(BaseModel):
+    session: PlaygroundSessionResponse
+    messages: list[PlaygroundMessageRecord] = Field(default_factory=list)
+    active_message_id: str | None = None
+
+
+class PlaygroundMessageStopRequest(BaseModel):
+    session_key: str = Field(default="default", min_length=1, max_length=100)
+    message_id: str = Field(..., min_length=1, max_length=64)
+
+
+class PlaygroundMessageResendRequest(BaseModel):
+    session_key: str = Field(default="default", min_length=1, max_length=100)
+    user_message_id: str = Field(..., min_length=1, max_length=64)
+
+
+class PlaygroundMessageEditRequest(BaseModel):
+    session_key: str = Field(default="default", min_length=1, max_length=100)
+    user_message_id: str = Field(..., min_length=1, max_length=64)
+    content: str = Field(..., min_length=1)

backend/app/schemas/alert.py

@@ -0,0 +1,5 @@
+from pydantic import BaseModel, Field
+
+
+class AlertResolutionRequest(BaseModel):
+    resolution: str = Field(..., min_length=1, max_length=1000)

backend/app/services/alert_ai_brief.py

@@ -0,0 +1,103 @@
+from __future__ import annotations
+
+from collections import Counter
+from typing import Any
+
+from sqlalchemy import func, select
+from sqlalchemy.ext.asyncio import AsyncSession
+
+from app.models.alert import Alert, AlertSeverity, AlertStatus
+from app.schemas.ai import AlertBriefRequest, SituationalAnalysisRequest
+
+
+def _format_counter(counter: Counter[str], empty_text: str = "无") -> str:
+    if not counter:
+        return empty_text
+    return "，".join(f"{key} {value}" for key, value in counter.items())
+
+
+async def build_alert_brief_request(
+    db: AsyncSession,
+    *,
+    alert_limit: int = 8,
+) -> tuple[SituationalAnalysisRequest, list[str], dict[str, Any]]:
+    recent_alerts_result = await db.execute(
+        select(Alert)
+        .order_by(Alert.created_at.desc(), Alert.id.desc())
+        .limit(max(alert_limit, 1))
+    )
+    total_result = await db.execute(select(func.count(Alert.id)))
+    active_result = await db.execute(select(func.count(Alert.id)).where(Alert.status == AlertStatus.ACTIVE))
+    acknowledged_result = await db.execute(
+        select(func.count(Alert.id)).where(Alert.status == AlertStatus.ACKNOWLEDGED)
+    )
+    resolved_result = await db.execute(select(func.count(Alert.id)).where(Alert.status == AlertStatus.RESOLVED))
+
+    recent_alerts = recent_alerts_result.scalars().all()
+    total_alerts = total_result.scalar() or 0
+    active_alerts = active_result.scalar() or 0
+    acknowledged_alerts = acknowledged_result.scalar() or 0
+    resolved_alerts = resolved_result.scalar() or 0
+
+    severity_counts = Counter((item.severity.value if item.severity else "unknown") for item in recent_alerts)
+    status_counts = Counter((item.status.value if item.status else "unknown") for item in recent_alerts)
+    datasource_counts = Counter((item.datasource_name or "未命名数据源") for item in recent_alerts)
+    active_datasource_counts = Counter(
+        (item.datasource_name or "未命名数据源")
+        for item in recent_alerts
+        if item.status == AlertStatus.ACTIVE
+    )
+
+    facts = [
+        f"告警总量 {total_alerts} 条，其中 active {active_alerts} 条、acknowledged {acknowledged_alerts} 条、resolved {resolved_alerts} 条。",
+        f"最近告警严重度分布：{_format_counter(severity_counts)}。",
+        f"最近告警状态分布：{_format_counter(status_counts)}。",
+        f"最近告警数据源分布：{_format_counter(Counter(dict(datasource_counts.most_common(6))))}。",
+    ]
+
+    if active_datasource_counts:
+        facts.append(
+            "当前待处理告警主要集中在："
+            + _format_counter(Counter(dict(active_datasource_counts.most_common(5))))
+            + "。"
+        )
+
+    if recent_alerts:
+        facts.append(
+            "最近告警摘录："
+            + "；".join(
+                [
+                    f"{item.datasource_name or '未命名数据源'} / {item.severity.value if item.severity else '-'} / {item.status.value if item.status else '-'} / {item.message or '-'}"
+                    for item in recent_alerts[:6]
+                ]
+            )
+        )
+
+    context = {
+        "source": "alerts",
+        "total_alerts": total_alerts,
+        "active_alerts": active_alerts,
+        "acknowledged_alerts": acknowledged_alerts,
+        "resolved_alerts": resolved_alerts,
+        "severity_distribution": dict(severity_counts),
+        "status_distribution": dict(status_counts),
+        "top_datasources": dict(datasource_counts.most_common(6)),
+        "top_active_datasources": dict(active_datasource_counts.most_common(5)),
+    }
+
+    return (
+        SituationalAnalysisRequest(
+            title="告警态势 AI 简报",
+            objective="基于当前告警总量、严重度、状态、数据源分布与最近告警摘录，生成一份面向值班人员的简明告警态势简报，突出待处理风险、告警集中点和优先动作。",
+            observations=facts,
+            constraints=[
+                "明确区分事实、推断与建议。",
+                "优先指出仍处于 active 状态且高严重度的告警簇。",
+                "不要把 acknowledged 或 resolved 告警误判成当前仍在扩大。",
+                "如果证据不足，请明确指出缺失的上下文。",
+            ],
+            context=context,
+        ),
+        facts,
+        context,
+    )

backend/app/services/bgp_ai_brief.py

@@ -0,0 +1,259 @@
+from __future__ import annotations
+
+from collections import Counter
+from typing import Any
+
+from sqlalchemy import func, select
+from sqlalchemy.ext.asyncio import AsyncSession
+
+from app.api.v1.bgp import BGP_SOURCES
+from app.models.bgp_anomaly import BGPAnomaly
+from app.models.bgp_incident import BGPIncident
+from app.models.bgp_observation import BGPObservation
+from app.schemas.ai import SituationalAnalysisRequest
+from app.services.bgp_collectors import build_bgp_collector_coverage
+from app.services.bgp_enrichment import lookup_prefix_geography
+
+
+def _format_counter(counter: dict[str, int], empty_text: str = "无") -> str:
+    if not counter:
+        return empty_text
+    return "，".join(f"{key} {value}" for key, value in counter.items())
+
+
+def _severity_rank(value: str | None) -> int:
+    order = {
+        "critical": 0,
+        "high": 1,
+        "medium": 2,
+        "low": 3,
+        "info": 4,
+    }
+    return order.get((value or "").lower(), 99)
+
+
+def _normalize_geo_key(country: str | None, city: str | None) -> str:
+    if city and country:
+        return f"{city}, {country}"
+    return city or country or "未知区域"
+
+
+def _top_counter_items(counter: Counter[str], limit: int = 5) -> dict[str, int]:
+    return {name: count for name, count in counter.most_common(limit) if name}
+
+
+def _collect_incident_regions(incidents: list[BGPIncident]) -> Counter[str]:
+    counter: Counter[str] = Counter()
+    for item in incidents:
+        for region in item.affected_regions or []:
+            if not isinstance(region, dict):
+                continue
+            counter[_normalize_geo_key(region.get("country"), region.get("city"))] += 1
+    return counter
+
+
+def _collect_collector_regions(collectors: list[dict[str, Any]]) -> Counter[str]:
+    counter: Counter[str] = Counter()
+    for item in collectors:
+        counter[_normalize_geo_key(item.get("country"), item.get("city"))] += int(item.get("recent_24h_observation_count") or 0)
+    return counter
+
+
+def _format_geo_evidence(prefix_geographies: dict[str, dict[str, Any]], limit: int = 6) -> str:
+    if not prefix_geographies:
+        return "没有命中 prefix geography 证据。"
+
+    rows = []
+    for prefix, item in list(prefix_geographies.items())[:limit]:
+        region = _normalize_geo_key(item.get("country"), item.get("city"))
+        source = item.get("source") or item.get("geography_mode") or "unknown"
+        as_hint = item.get("asn")
+        as_name = item.get("as_name")
+        as_text = ""
+        if as_hint:
+            as_text = f" / ASN AS{as_hint}"
+            if as_name:
+                as_text += f" ({as_name})"
+        rows.append(f"{prefix} -> {region} / 来源 {source}{as_text}")
+    return "；".join(rows)
+
+
+async def build_bgp_brief_request(
+    db: AsyncSession,
+    *,
+    incident_limit: int = 5,
+    anomaly_limit: int = 6,
+    collector_limit: int = 5,
+) -> tuple[SituationalAnalysisRequest, list[str], dict[str, int | str | dict[str, int]]]:
+    incidents_result = await db.execute(
+        select(BGPIncident)
+        .order_by(BGPIncident.created_at.desc(), BGPIncident.id.desc())
+        .limit(max(incident_limit, 1))
+    )
+    anomalies_result = await db.execute(
+        select(BGPAnomaly)
+        .order_by(BGPAnomaly.created_at.desc(), BGPAnomaly.id.desc())
+        .limit(max(anomaly_limit, 1))
+    )
+    observations_result = await db.execute(
+        select(BGPObservation).where(BGPObservation.source.in_(BGP_SOURCES))
+    )
+    incident_count_result = await db.execute(select(func.count(BGPIncident.id)))
+    anomaly_count_result = await db.execute(select(func.count(BGPAnomaly.id)))
+
+    incidents = incidents_result.scalars().all()
+    anomalies = anomalies_result.scalars().all()
+    observations = observations_result.scalars().all()
+    collectors = await build_bgp_collector_coverage(db, source_filter=BGP_SOURCES)
+
+    total_incidents = incident_count_result.scalar() or 0
+    total_anomalies = anomaly_count_result.scalar() or 0
+    total_observations = len(observations)
+    active_collectors = [item for item in collectors if item["observation_count"] > 0]
+
+    incident_status_counts = Counter((item.status or "unknown") for item in incidents)
+    incident_severity_counts = Counter((item.severity or "unknown") for item in incidents)
+    incident_type_counts = Counter((item.incident_type or "unknown") for item in incidents)
+    anomaly_type_counts = Counter((item.anomaly_type or "unknown") for item in anomalies)
+    event_type_counts = Counter((item.event_type or "unknown") for item in observations)
+    incident_region_counts = _collect_incident_regions(incidents)
+
+    top_collectors = sorted(
+        active_collectors,
+        key=lambda item: (
+            -int(item["recent_24h_observation_count"]),
+            -int(item["observation_count"]),
+            str(item["collector"]),
+        ),
+    )[: max(collector_limit, 1)]
+    collector_region_counts = _collect_collector_regions(top_collectors)
+
+    prefix_candidates = sorted(
+        {
+            prefix
+            for item in incidents
+            for prefix in (item.affected_prefixes or [])
+            if prefix
+        }
+        | {item.prefix for item in anomalies if item.prefix}
+    )
+    prefix_geographies = await lookup_prefix_geography(db, prefix_candidates) if prefix_candidates else {}
+    geography_region_counts = Counter(
+        _normalize_geo_key(item.get("country"), item.get("city"))
+        for item in prefix_geographies.values()
+        if item.get("country") or item.get("city")
+    )
+    hotspot_region_counts = geography_region_counts + incident_region_counts
+    collector_bias_regions = [
+        region
+        for region, count in collector_region_counts.most_common(3)
+        if count > hotspot_region_counts.get(region, 0)
+    ]
+
+    observations_lines: list[str] = [
+        f"当前共有 {total_incidents} 起 BGP incidents、{total_anomalies} 条 anomalies、{total_observations} 条原始观测事件。",
+        f"活跃观测站 {len(active_collectors)} 个；近 24 小时事件数合计 {sum(int(item['recent_24h_observation_count']) for item in active_collectors)}。",
+        f"最近 incidents 严重度分布：{_format_counter(dict(sorted(incident_severity_counts.items(), key=lambda item: _severity_rank(item[0]))))}。",
+        f"最近 incidents 状态分布：{_format_counter(dict(incident_status_counts))}。",
+        f"最近 incidents 类型分布：{_format_counter(dict(incident_type_counts.most_common(5)))}。",
+        f"最近 anomalies 类型分布：{_format_counter(dict(anomaly_type_counts.most_common(6)))}。",
+        f"观测事件类型分布：{_format_counter(dict(event_type_counts.most_common(6)))}。",
+    ]
+
+    if hotspot_region_counts:
+        observations_lines.append(
+            "区域热点事实层："
+            + _format_counter(_top_counter_items(hotspot_region_counts, limit=5), empty_text="无明显区域聚集")
+            + "。"
+        )
+
+    if prefix_geographies:
+        observations_lines.append("Prefix geography 证据：" + _format_geo_evidence(prefix_geographies))
+
+    if collector_bias_regions:
+        observations_lines.append(
+            "观测偏差提示：重点观测站最近 24h 活跃度更集中在 "
+            + "、".join(collector_bias_regions)
+            + "，这些区域的事件升温结论需要结合 prefix geography 与 affected regions 交叉验证。"
+        )
+    elif top_collectors:
+        observations_lines.append(
+            "观测偏差提示：当前未发现明显高于区域热点事实层的单一观测站集中区域，但仍需区分 collector coverage 与真实区域风险。"
+        )
+
+    if incidents:
+        observations_lines.append(
+            "最近 incident 摘要：" + "；".join(
+                [
+                    f"{item.incident_type} / {item.severity} / {item.status}"
+                    f" / 前缀 {', '.join(item.affected_prefixes[:2]) if item.affected_prefixes else '-'}"
+                    f" / 观测站 {len(item.affected_collectors or [])} 个"
+                    for item in incidents
+                ]
+            )
+        )
+
+    if anomalies:
+        observations_lines.append(
+            "最近 anomaly 摘要：" + "；".join(
+                [
+                    f"{item.anomaly_type} / {item.severity}"
+                    f" / 前缀 {item.prefix or '-'}"
+                    f" / ASN {item.new_origin_asn or item.origin_asn or '-'}"
+                    for item in anomalies
+                ]
+            )
+        )
+
+    if top_collectors:
+        observations_lines.append(
+            "重点观测站：" + "；".join(
+                [
+                    f"{item['collector']} ({', '.join([part for part in [item.get('city'), item.get('country')] if part]) or '未知位置'})"
+                    f" / 近24h {item['recent_24h_observation_count']} 条"
+                    f" / 前缀 {item['prefix_count']} 个"
+                    for item in top_collectors
+                ]
+            )
+        )
+
+    context = {
+        "source": "bgp-overview",
+        "incident_total": total_incidents,
+        "anomaly_total": total_anomalies,
+        "observation_total": total_observations,
+        "active_collectors": len(active_collectors),
+        "top_incident_types": dict(incident_type_counts.most_common(5)),
+        "top_anomaly_types": dict(anomaly_type_counts.most_common(6)),
+        "top_event_types": dict(event_type_counts.most_common(6)),
+        "region_hotspots": _top_counter_items(hotspot_region_counts, limit=6),
+        "incident_regions": _top_counter_items(incident_region_counts, limit=6),
+        "collector_bias_regions": collector_bias_regions,
+        "prefix_geography_sources": dict(
+            Counter(str(item.get("source") or "unknown") for item in prefix_geographies.values()).most_common(5)
+        ),
+        "prefix_geography_sample": {
+            prefix: {
+                "country": item.get("country"),
+                "city": item.get("city"),
+                "source": item.get("source"),
+                "asn": item.get("asn"),
+                "as_name": item.get("as_name"),
+            }
+            for prefix, item in list(prefix_geographies.items())[:8]
+        },
+    }
+
+    return SituationalAnalysisRequest(
+        title="BGP 态势 AI 简报",
+        objective="基于当前 BGP incidents、anomalies、原始观测事件、观测站覆盖与 prefix geography 证据，生成一份面向操作员的简明态势简报，突出区域热点、观测偏差、当前风险、证据和优先动作。",
+        observations=observations_lines,
+        constraints=[
+            "明确区分事实、推断与建议。",
+            "优先指出需要立即关注的高严重度 incident 或异常模式。",
+            "需要单独指出哪些区域结论来自 prefix geography / affected regions，哪些可能受 collector coverage 偏差影响。",
+            "结论应服务值班排障，不要写成泛泛的模型演示文案。",
+            "如果证据不足，要明确指出缺失数据。",
+        ],
+        context=context,
+    ), observations_lines, context

backend/app/services/bgp_ai_brief_store.py

@@ -0,0 +1,160 @@
+from __future__ import annotations
+
+import json
+from dataclasses import dataclass
+from datetime import UTC, datetime
+from pathlib import Path
+from typing import Any
+from uuid import uuid4
+
+from app.core.config import ROOT_DIR
+from app.schemas.ai import BGPBriefRecordResponse, BGPBriefRecordSummary, SituationalAnalysisResponse
+
+
+_BRIEF_STORAGE_DIR = ROOT_DIR / "data" / "ai" / "bgp-briefs"
+_METADATA_PREFIX = "<!-- planet-bgp-brief-meta "
+_METADATA_SUFFIX = " -->"
+_BRIEF_TITLE = "BGP AI 简报"
+
+
+@dataclass(slots=True)
+class _StoredBrief:
+    id: str
+    title: str
+    provider: str
+    model: str
+    request_id: str | None
+    generated_at: str
+    content_markdown: str
+    facts: list[str]
+    context: dict[str, Any]
+    path: Path
+
+
+def _ensure_storage_dir() -> Path:
+    _BRIEF_STORAGE_DIR.mkdir(parents=True, exist_ok=True)
+    return _BRIEF_STORAGE_DIR
+
+
+def _build_metadata_line(metadata: dict[str, Any]) -> str:
+    return f"{_METADATA_PREFIX}{json.dumps(metadata, ensure_ascii=False)}{_METADATA_SUFFIX}"
+
+
+def _parse_brief_file(path: Path) -> _StoredBrief | None:
+    try:
+        raw_text = path.read_text(encoding="utf-8")
+    except OSError:
+        return None
+
+    first_line, separator, remainder = raw_text.partition("\n")
+    if not separator or not first_line.startswith(_METADATA_PREFIX) or not first_line.endswith(_METADATA_SUFFIX):
+        return None
+
+    metadata_payload = first_line[len(_METADATA_PREFIX) : -len(_METADATA_SUFFIX)]
+
+    try:
+        metadata = json.loads(metadata_payload)
+    except json.JSONDecodeError:
+        return None
+
+    return _StoredBrief(
+        id=str(metadata.get("id") or path.stem),
+        title=str(metadata.get("title") or _BRIEF_TITLE),
+        provider=str(metadata.get("provider") or "-"),
+        model=str(metadata.get("model") or "-"),
+        request_id=metadata.get("request_id"),
+        generated_at=str(metadata.get("generated_at") or datetime.fromtimestamp(path.stat().st_mtime, UTC).isoformat()),
+        content_markdown=remainder.lstrip("\n"),
+        facts=list(metadata.get("facts") or []),
+        context=dict(metadata.get("context") or {}),
+        path=path,
+    )
+
+
+def list_bgp_brief_records(limit: int = 50) -> list[BGPBriefRecordSummary]:
+    storage_dir = _ensure_storage_dir()
+    records: list[_StoredBrief] = []
+
+    for path in storage_dir.glob("*.md"):
+        parsed = _parse_brief_file(path)
+        if parsed is not None:
+            records.append(parsed)
+
+    records.sort(key=lambda item: item.generated_at, reverse=True)
+
+    return [
+        BGPBriefRecordSummary(
+            id=item.id,
+            title=item.title,
+            provider=item.provider,
+            model=item.model,
+            request_id=item.request_id,
+            generated_at=item.generated_at,
+        )
+        for item in records[: max(limit, 1)]
+    ]
+
+
+def get_bgp_brief_record(brief_id: str) -> BGPBriefRecordResponse | None:
+    path = _ensure_storage_dir() / f"{brief_id}.md"
+    parsed = _parse_brief_file(path)
+    if parsed is None:
+        return None
+
+    return BGPBriefRecordResponse(
+        id=parsed.id,
+        title=parsed.title,
+        provider=parsed.provider,
+        model=parsed.model,
+        request_id=parsed.request_id,
+        generated_at=parsed.generated_at,
+        content_markdown=parsed.content_markdown,
+        facts=parsed.facts,
+        context=parsed.context,
+    )
+
+
+def get_latest_bgp_brief_record() -> BGPBriefRecordResponse | None:
+    summaries = list_bgp_brief_records(limit=1)
+    if not summaries:
+        return None
+    return get_bgp_brief_record(summaries[0].id)
+
+
+def save_bgp_brief_record(
+    analysis: SituationalAnalysisResponse,
+    *,
+    request_id: str | None,
+    facts: list[str] | None = None,
+    context: dict[str, Any] | None = None,
+    generated_at: datetime | None = None,
+) -> BGPBriefRecordResponse:
+    created_at = generated_at or datetime.now(UTC)
+    brief_id = f"{created_at.strftime('%Y%m%dT%H%M%SZ')}-{uuid4().hex[:8]}"
+    path = _ensure_storage_dir() / f"{brief_id}.md"
+
+    metadata = {
+        "id": brief_id,
+        "title": _BRIEF_TITLE,
+        "provider": analysis.provider,
+        "model": analysis.model,
+        "request_id": request_id,
+        "generated_at": created_at.isoformat(),
+        "facts": facts or [],
+        "context": context or {},
+    }
+
+    markdown_text = f"{_build_metadata_line(metadata)}\n\n{analysis.content.rstrip()}\n"
+    path.write_text(markdown_text, encoding="utf-8")
+
+    return BGPBriefRecordResponse(
+        id=brief_id,
+        title=_BRIEF_TITLE,
+        provider=analysis.provider,
+        model=analysis.model,
+        request_id=request_id,
+        generated_at=created_at.isoformat(),
+        content_markdown=analysis.content,
+        facts=facts or [],
+        context=context or {},
+    )

backend/app/services/bgp_collectors.py

@@ -6,7 +6,7 @@ from collections import defaultdict
 from datetime import UTC, datetime, timedelta
 from typing import Any
 
-from sqlalchemy import select
+from sqlalchemy import case, distinct, func, select
 from sqlalchemy.ext.asyncio import AsyncSession
 
 from app.core.time import to_iso8601_utc
@@ -14,6 +14,16 @@ from app.models.bgp_observation import BGPObservation
 from app.services.collectors.bgp_common import RIPE_RIS_COLLECTOR_COORDS
 
 
+def _collector_base_filters(source_filter: tuple[str, ...] | None) -> list[Any]:
+    filters: list[Any] = [
+        BGPObservation.collector.isnot(None),
+        func.length(func.btrim(BGPObservation.collector)) > 0,
+    ]
+    if source_filter:
+        filters.append(BGPObservation.source.in_(source_filter))
+    return filters
+
+
 async def build_bgp_collector_coverage(
     db: AsyncSession,
     *,
@@ -24,88 +34,148 @@ async def build_bgp_collector_coverage(
     recent_24h_threshold = now - timedelta(hours=24)
     recent_7d_threshold = now - timedelta(days=7)
 
-    stmt = select(BGPObservation).order_by(BGPObservation.observed_at.desc(), BGPObservation.id.desc())
-    if source_filter:
-        stmt = stmt.where(BGPObservation.source.in_(source_filter))
+    filters = _collector_base_filters(source_filter)
+    country_expr = func.nullif(BGPObservation.collector_geo["country"].as_string(), "")
+    city_expr = func.nullif(BGPObservation.collector_geo["city"].as_string(), "")
 
-    result = await db.execute(stmt)
-    records = list(result.scalars().all())
+    aggregate_stmt = (
+        select(
+            BGPObservation.collector.label("collector"),
+            func.count(BGPObservation.id).label("observation_count"),
+            func.count(distinct(BGPObservation.prefix)).label("prefix_count"),
+            func.count(distinct(BGPObservation.origin_asn)).label("origin_asn_count"),
+            func.count(distinct(BGPObservation.peer_asn)).label("peer_asn_count"),
+            func.sum(case((BGPObservation.observed_at >= recent_15m_threshold, 1), else_=0)).label("recent_15m_observation_count"),
+            func.sum(case((BGPObservation.observed_at >= recent_24h_threshold, 1), else_=0)).label("recent_24h_observation_count"),
+            func.sum(case((BGPObservation.observed_at >= recent_7d_threshold, 1), else_=0)).label("recent_7d_observation_count"),
+            func.count(distinct(case((BGPObservation.observed_at >= recent_15m_threshold, BGPObservation.prefix), else_=None))).label("recent_15m_prefix_count"),
+            func.count(distinct(case((BGPObservation.observed_at >= recent_24h_threshold, BGPObservation.prefix), else_=None))).label("recent_24h_prefix_count"),
+            func.count(distinct(case((BGPObservation.observed_at >= recent_7d_threshold, BGPObservation.prefix), else_=None))).label("recent_7d_prefix_count"),
+            func.max(BGPObservation.observed_at).label("latest_observed_at"),
+        )
+        .where(*filters)
+        .group_by(BGPObservation.collector)
+    )
+    aggregate_rows = (await db.execute(aggregate_stmt)).all()
+
+    latest_subquery = (
+        select(
+            BGPObservation.collector.label("collector"),
+            BGPObservation.event_type.label("latest_event_type"),
+            country_expr.label("country"),
+            city_expr.label("city"),
+            func.row_number()
+            .over(
+                partition_by=BGPObservation.collector,
+                order_by=(BGPObservation.observed_at.desc(), BGPObservation.id.desc()),
+            )
+            .label("rn"),
+        )
+        .where(*filters)
+        .subquery()
+    )
+    latest_rows = (
+        await db.execute(
+            select(
+                latest_subquery.c.collector,
+                latest_subquery.c.latest_event_type,
+                latest_subquery.c.country,
+                latest_subquery.c.city,
+            ).where(latest_subquery.c.rn == 1)
+        )
+    ).all()
+
+    event_counts_subquery = (
+        select(
+            BGPObservation.collector.label("collector"),
+            BGPObservation.event_type.label("event_type"),
+            func.count(BGPObservation.id).label("count"),
+            func.row_number()
+            .over(
+                partition_by=BGPObservation.collector,
+                order_by=(func.count(BGPObservation.id).desc(), BGPObservation.event_type.asc()),
+            )
+            .label("rn"),
+        )
+        .where(*filters)
+        .group_by(BGPObservation.collector, BGPObservation.event_type)
+        .subquery()
+    )
+    top_event_rows = (
+        await db.execute(
+            select(
+                event_counts_subquery.c.collector,
+                event_counts_subquery.c.event_type,
+                event_counts_subquery.c.count,
+            ).where(event_counts_subquery.c.rn <= 3)
+        )
+    ).all()
+
+    scope_rows = (
+        await db.execute(
+            select(
+                BGPObservation.collector.label("collector"),
+                country_expr.label("country"),
+                city_expr.label("city"),
+            )
+            .where(*filters)
+            .distinct()
+        )
+    ).all()
+
+    latest_by_collector = {
+        row.collector: {
+            "latest_event_type": row.latest_event_type,
+            "country": row.country,
+            "city": row.city,
+        }
+        for row in latest_rows
+    }
+
+    scope_by_collector: dict[str, dict[str, set[str]]] = defaultdict(lambda: {"countries": set(), "cities": set()})
+    for row in scope_rows:
+        if row.country:
+            scope_by_collector[row.collector]["countries"].add(row.country)
+        if row.city:
+            scope_by_collector[row.collector]["cities"].add(row.city)
+
+    top_events_by_collector: dict[str, list[dict[str, Any]]] = defaultdict(list)
+    for row in top_event_rows:
+        top_events_by_collector[row.collector].append(
+            {"event_type": row.event_type, "count": row.count}
+        )
 
     by_collector: dict[str, dict[str, Any]] = {}
-    for record in records:
-        collector = str(record.collector or "").strip()
-        if not collector:
-            continue
+    for row in aggregate_rows:
+        collector = row.collector
+        latest = latest_by_collector.get(collector, {})
+        fallback_location = RIPE_RIS_COLLECTOR_COORDS.get(collector, {})
+        scope = scope_by_collector.get(collector, {"countries": set(), "cities": set()})
 
-        coverage = by_collector.get(collector)
-        if coverage is None:
-            location = record.collector_geo or RIPE_RIS_COLLECTOR_COORDS.get(collector, {})
-            coverage = {
-                "collector": collector,
-                "city": location.get("city"),
-                "country": location.get("country"),
-                "latitude": location.get("latitude"),
-                "longitude": location.get("longitude"),
-                "observation_count": 0,
-                "prefixes": set(),
-                "origin_asns": set(),
-                "peer_asns": set(),
-                "event_types": defaultdict(int),
-                "countries": set(),
-                "cities": set(),
-                "recent_15m_observation_count": 0,
-                "recent_24h_observation_count": 0,
-                "recent_7d_observation_count": 0,
-                "recent_15m_prefixes": set(),
-                "recent_24h_prefixes": set(),
-                "recent_7d_prefixes": set(),
-                "latest_observed_at": None,
-                "latest_event_type": None,
-            }
-            by_collector[collector] = coverage
-
-        coverage["observation_count"] += 1
-        if record.prefix:
-            coverage["prefixes"].add(record.prefix)
-        if record.origin_asn is not None:
-            coverage["origin_asns"].add(record.origin_asn)
-        if record.peer_asn is not None:
-            coverage["peer_asns"].add(record.peer_asn)
-        if record.event_type:
-            coverage["event_types"][record.event_type] += 1
-
-        observed_at = record.observed_at
-        if observed_at is not None:
-            aware_observed_at = (
-                observed_at.astimezone(UTC)
-                if observed_at.tzinfo
-                else observed_at.replace(tzinfo=UTC)
-            )
-            if aware_observed_at >= recent_15m_threshold:
-                coverage["recent_15m_observation_count"] += 1
-                if record.prefix:
-                    coverage["recent_15m_prefixes"].add(record.prefix)
-            if aware_observed_at >= recent_24h_threshold:
-                coverage["recent_24h_observation_count"] += 1
-                if record.prefix:
-                    coverage["recent_24h_prefixes"].add(record.prefix)
-            if aware_observed_at >= recent_7d_threshold:
-                coverage["recent_7d_observation_count"] += 1
-                if record.prefix:
-                    coverage["recent_7d_prefixes"].add(record.prefix)
-
-        geo = record.collector_geo or {}
-        if geo.get("country"):
-            coverage["countries"].add(geo["country"])
-        if geo.get("city"):
-            coverage["cities"].add(geo["city"])
-
-        current_latest = coverage["latest_observed_at"]
-        if current_latest is None or (
-            record.observed_at is not None and record.observed_at > current_latest
-        ):
-            coverage["latest_observed_at"] = record.observed_at
-            coverage["latest_event_type"] = record.event_type
+        by_collector[collector] = {
+            "collector": collector,
+            "city": latest.get("city") or fallback_location.get("city"),
+            "country": latest.get("country") or fallback_location.get("country"),
+            "latitude": fallback_location.get("latitude"),
+            "longitude": fallback_location.get("longitude"),
+            "observation_count": row.observation_count or 0,
+            "prefix_count": row.prefix_count or 0,
+            "origin_asn_count": row.origin_asn_count or 0,
+            "peer_asn_count": row.peer_asn_count or 0,
+            "recent_15m_observation_count": row.recent_15m_observation_count or 0,
+            "recent_24h_observation_count": row.recent_24h_observation_count or 0,
+            "recent_7d_observation_count": row.recent_7d_observation_count or 0,
+            "recent_15m_prefix_count": row.recent_15m_prefix_count or 0,
+            "recent_24h_prefix_count": row.recent_24h_prefix_count or 0,
+            "recent_7d_prefix_count": row.recent_7d_prefix_count or 0,
+            "top_event_types": top_events_by_collector.get(collector, []),
+            "latest_observed_at": to_iso8601_utc(row.latest_observed_at),
+            "latest_event_type": latest.get("latest_event_type"),
+            "baseline_scope": {
+                "countries": sorted(scope["countries"]),
+                "cities": sorted(scope["cities"]),
+            },
+        }
 
     for collector, location in RIPE_RIS_COLLECTOR_COORDS.items():
         if collector in by_collector:
@@ -117,57 +187,22 @@ async def build_bgp_collector_coverage(
             "latitude": location.get("latitude"),
             "longitude": location.get("longitude"),
             "observation_count": 0,
-            "prefixes": set(),
-            "origin_asns": set(),
-            "peer_asns": set(),
-            "event_types": defaultdict(int),
-            "countries": {location.get("country")} if location.get("country") else set(),
-            "cities": {location.get("city")} if location.get("city") else set(),
+            "prefix_count": 0,
+            "origin_asn_count": 0,
+            "peer_asn_count": 0,
             "recent_15m_observation_count": 0,
             "recent_24h_observation_count": 0,
             "recent_7d_observation_count": 0,
-            "recent_15m_prefixes": set(),
-            "recent_24h_prefixes": set(),
-            "recent_7d_prefixes": set(),
+            "recent_15m_prefix_count": 0,
+            "recent_24h_prefix_count": 0,
+            "recent_7d_prefix_count": 0,
+            "top_event_types": [],
             "latest_observed_at": None,
             "latest_event_type": None,
+            "baseline_scope": {
+                "countries": [location["country"]] if location.get("country") else [],
+                "cities": [location["city"]] if location.get("city") else [],
+            },
         }
 
-    results: list[dict[str, Any]] = []
-    for collector in sorted(by_collector.keys()):
-        item = by_collector[collector]
-        top_event_types = sorted(
-            item["event_types"].items(),
-            key=lambda pair: (-pair[1], pair[0]),
-        )
-        results.append(
-            {
-                "collector": item["collector"],
-                "city": item["city"],
-                "country": item["country"],
-                "latitude": item["latitude"],
-                "longitude": item["longitude"],
-                "observation_count": item["observation_count"],
-                "prefix_count": len(item["prefixes"]),
-                "origin_asn_count": len(item["origin_asns"]),
-                "peer_asn_count": len(item["peer_asns"]),
-                "recent_15m_observation_count": item["recent_15m_observation_count"],
-                "recent_24h_observation_count": item["recent_24h_observation_count"],
-                "recent_7d_observation_count": item["recent_7d_observation_count"],
-                "recent_15m_prefix_count": len(item["recent_15m_prefixes"]),
-                "recent_24h_prefix_count": len(item["recent_24h_prefixes"]),
-                "recent_7d_prefix_count": len(item["recent_7d_prefixes"]),
-                "top_event_types": [
-                    {"event_type": event_type, "count": count}
-                    for event_type, count in top_event_types[:3]
-                ],
-                "latest_observed_at": to_iso8601_utc(item["latest_observed_at"]),
-                "latest_event_type": item["latest_event_type"],
-                "baseline_scope": {
-                    "countries": sorted(country for country in item["countries"] if country),
-                    "cities": sorted(city for city in item["cities"] if city),
-                },
-            }
-        )
-
-    return results
+    return [by_collector[collector] for collector in sorted(by_collector.keys())]

backend/app/services/bgp_enrichment.py

@@ -7,7 +7,7 @@ from collections import defaultdict
 from datetime import UTC, datetime
 from typing import Any
 
-from sqlalchemy import select, text
+from sqlalchemy import Integer, cast, select, text
 from sqlalchemy.ext.asyncio import AsyncSession
 
 from app.core.countries import get_country_centroid, normalize_country
@@ -231,6 +231,13 @@ async def _lookup_prefix_geography(
     return results
 
 
+async def lookup_prefix_geography(
+    db: AsyncSession,
+    prefix_values: list[str],
+) -> dict[str, dict[str, Any]]:
+    return await _lookup_prefix_geography(db, prefix_values)
+
+
 async def enrich_bgp_events_for_batch(
     db: AsyncSession,
     *,
@@ -261,29 +268,40 @@ async def enrich_bgp_events_for_batch(
     historical_prefix_baseline: dict[str, dict[str, Any]] = {}
     if prefix_values:
         previous_result = await db.execute(
-            select(BGPObservation).where(
+            select(
+                BGPObservation.prefix,
+                BGPObservation.origin_asn,
+                BGPObservation.collector,
+                BGPObservation.collector_geo,
+            ).where(
                 BGPObservation.source == source,
                 BGPObservation.prefix.in_(prefix_values),
             )
         )
-        by_prefix: defaultdict[str, list[BGPObservation]] = defaultdict(list)
-        for observation in previous_result.scalars().all():
-            if observation.prefix:
-                by_prefix[observation.prefix].append(observation)
+        by_prefix: defaultdict[str, list[dict[str, Any]]] = defaultdict(list)
+        for prefix, origin_asn, collector, collector_geo in previous_result.all():
+            if prefix:
+                by_prefix[str(prefix)].append(
+                    {
+                        "origin_asn": origin_asn,
+                        "collector": collector,
+                        "collector_geo": collector_geo or {},
+                    }
+                )
 
         for prefix, observations in by_prefix.items():
             unique_origins = sorted(
                 {
-                    observation.origin_asn
+                    observation["origin_asn"]
                     for observation in observations
-                    if observation.origin_asn is not None
+                    if observation["origin_asn"] is not None
                 }
             )
             unique_collectors = sorted(
                 {
-                    observation.collector
+                    observation["collector"]
                     for observation in observations
-                    if observation.collector
+                    if observation["collector"]
                 }
             )
             historical_prefix_baseline[prefix] = {
@@ -292,9 +310,9 @@ async def enrich_bgp_events_for_batch(
                 "historical_observation_count": len(observations),
                 "historical_regions": _compact_locations(
                     [
-                        observation.collector_geo or {}
+                        observation["collector_geo"] or {}
                         for observation in observations
-                        if observation.collector_geo
+                        if observation["collector_geo"]
                     ]
                 ),
             }
@@ -303,7 +321,13 @@ async def enrich_bgp_events_for_batch(
     prefix_geographies = await _lookup_prefix_geography(db, prefix_values) if prefix_values else {}
     if origin_asns:
         peeringdb_result = await db.execute(
-            select(CollectedData).where(CollectedData.source == "peeringdb_network")
+            select(CollectedData)
+            .where(CollectedData.source == "peeringdb_network")
+            .where(CollectedData.is_current.is_(True))
+            .where(
+                cast(CollectedData.extra_data["asn"].as_string(), Integer).in_(origin_asns),
+            )
+            .order_by(CollectedData.id.desc())
         )
         for record in peeringdb_result.scalars().all():
             metadata = record.extra_data or {}

backend/app/services/bgp_incidents.py

@@ -48,14 +48,36 @@ def _collector_regions_from_anomaly(anomaly: BGPAnomaly) -> list[dict]:
     return collected
 
 
-def _dedupe_collected_records(records: list[CollectedData]) -> list[CollectedData]:
-    latest_by_key: dict[str, CollectedData] = {}
-    for record in records:
-        dedupe_key = str(record.source_id or record.entity_key or record.name or record.id)
-        existing = latest_by_key.get(dedupe_key)
-        if existing is None or (record.id or 0) > (existing.id or 0):
-            latest_by_key[dedupe_key] = record
-    return list(latest_by_key.values())
+async def _load_current_infrastructure_records(
+    db: AsyncSession,
+) -> tuple[list[CollectedData], list[CollectedData], list[CollectedData]]:
+    result = await db.execute(
+        select(CollectedData)
+        .where(
+            CollectedData.source.in_(
+                (
+                    "arcgis_landing_points",
+                    "arcgis_cable_landing_relation",
+                    "arcgis_cables",
+                )
+            )
+        )
+        .where(CollectedData.is_current.is_(True))
+        .order_by(CollectedData.source.asc(), CollectedData.id.desc())
+    )
+    grouped_records = {
+        "arcgis_landing_points": [],
+        "arcgis_cable_landing_relation": [],
+        "arcgis_cables": [],
+    }
+    for record in result.scalars().all():
+        grouped_records.setdefault(record.source, []).append(record)
+
+    return (
+        grouped_records["arcgis_landing_points"],
+        grouped_records["arcgis_cable_landing_relation"],
+        grouped_records["arcgis_cables"],
+    )
 
 
 async def infer_related_infrastructure(
@@ -75,19 +97,9 @@ async def infer_related_infrastructure(
     if not valid_regions:
         return {"related_cables": [], "related_ixps": []}
 
-    landing_result = await db.execute(
-        select(CollectedData).where(CollectedData.source == "arcgis_landing_points")
+    landing_records, relation_records, cable_records = await _load_current_infrastructure_records(
+        db,
     )
-    relation_result = await db.execute(
-        select(CollectedData).where(CollectedData.source == "arcgis_cable_landing_relation")
-    )
-    cable_result = await db.execute(
-        select(CollectedData).where(CollectedData.source == "arcgis_cables")
-    )
-
-    landing_records = _dedupe_collected_records(list(landing_result.scalars().all()))
-    relation_records = _dedupe_collected_records(list(relation_result.scalars().all()))
-    cable_records = _dedupe_collected_records(list(cable_result.scalars().all()))
 
     city_to_cable_ids: dict[int, list[int]] = {}
     for relation in relation_records:

backend/app/services/collectors/__init__.py

@@ -35,6 +35,7 @@ from app.services.collectors.bgpstream import BGPStreamBackfillCollector
 from app.services.collectors.iptoasn import IPtoASNPrefixGeoCollector
 from app.services.collectors.opengeofeed import OpenGeoFeedPrefixGeoCollector
 from app.services.collectors.nro_delegated import NRODelegatedPrefixGeoCollector
+from app.services.collectors.news_live_streams import NewsLiveStreamsCollector
 
 collector_registry.register(TOP500Collector())
 collector_registry.register(EpochAIGPUCollector())
@@ -61,3 +62,4 @@ collector_registry.register(BGPStreamBackfillCollector())
 collector_registry.register(IPtoASNPrefixGeoCollector())
 collector_registry.register(OpenGeoFeedPrefixGeoCollector())
 collector_registry.register(NRODelegatedPrefixGeoCollector())
+collector_registry.register(NewsLiveStreamsCollector())

backend/app/services/collectors/base.py

@@ -1,5 +1,6 @@
 """Base collector class for all data sources"""
 
+import asyncio
 from abc import ABC, abstractmethod
 from typing import Dict, List, Any, Optional
 from datetime import UTC, datetime
@@ -166,6 +167,59 @@ class BaseCollector(ABC):
         await db.commit()
         return snapshot.id
 
+    async def _rollback_incomplete_run(
+        self,
+        db: AsyncSession,
+        *,
+        task_id: int,
+        snapshot_id: Optional[int],
+        reason: str,
+    ) -> None:
+        from app.models.collected_data import CollectedData
+        from app.models.data_snapshot import DataSnapshot
+
+        await db.execute(CollectedData.__table__.delete().where(CollectedData.task_id == task_id))
+
+        parent_snapshot_id: Optional[int] = None
+        if snapshot_id is not None:
+            snapshot = await db.get(DataSnapshot, snapshot_id)
+            if snapshot:
+                parent_snapshot_id = snapshot.parent_snapshot_id
+                snapshot.status = "cancelled"
+                snapshot.is_current = False
+                snapshot.completed_at = datetime.now(UTC)
+                summary = dict(snapshot.summary or {})
+                summary["rollback"] = True
+                summary["rollback_reason"] = reason
+                snapshot.summary = summary
+
+        await db.execute(
+            text(
+                """
+                UPDATE collected_data
+                SET is_current = FALSE
+                WHERE source = :source
+                """
+            ),
+            {"source": self.name},
+        )
+
+        if parent_snapshot_id is not None:
+            parent_snapshot = await db.get(DataSnapshot, parent_snapshot_id)
+            if parent_snapshot:
+                parent_snapshot.is_current = True
+
+            await db.execute(
+                text(
+                    """
+                    UPDATE collected_data
+                    SET is_current = TRUE
+                    WHERE snapshot_id = :snapshot_id
+                    """
+                ),
+                {"snapshot_id": parent_snapshot_id},
+            )
+
     async def run(self, db: AsyncSession) -> Dict[str, Any]:
         """Full pipeline: fetch -> transform -> save"""
         from app.services.collectors.registry import collector_registry
@@ -227,7 +281,24 @@ class BaseCollector(ABC):
                 "records_processed": records_count,
                 "execution_time_seconds": (datetime.now(UTC) - start_time).total_seconds(),
             }
+        except asyncio.CancelledError:
+            await db.rollback()
+            task.status = "cancelled"
+            task.phase = "cancelled"
+            task.error_message = "Collection cancelled by operator and rolled back"
+            task.completed_at = datetime.now(UTC)
+            if snapshot_id is not None:
+                await self._rollback_incomplete_run(
+                    db,
+                    task_id=task_id,
+                    snapshot_id=snapshot_id,
+                    reason="cancelled_by_operator",
+                )
+            await db.commit()
+            await self._publish_task_update(force=True)
+            raise
         except Exception as e:
+            await db.rollback()
             task.status = "failed"
             task.phase = "failed"
             task.error_message = str(e)
@@ -276,20 +347,34 @@ class BaseCollector(ABC):
         updated_count = 0
         unchanged_count = 0
         seen_entity_keys: set[str] = set()
-        previous_current_keys: set[str] = set()
+        progress_commit_interval = 1000
 
         previous_current_result = await db.execute(
-            select(CollectedData.entity_key).where(
+            select(CollectedData)
+            .where(
                 CollectedData.source == self.name,
                 CollectedData.is_current == True,
             )
+            .order_by(CollectedData.entity_key.asc(), CollectedData.collected_at.desc().nullslast(), CollectedData.id.desc())
         )
-        previous_current_keys = {row[0] for row in previous_current_result.fetchall() if row[0]}
+        previous_current_records = previous_current_result.scalars().all()
+        previous_current_keys = {record.entity_key for record in previous_current_records if record.entity_key}
+        previous_current_map: dict[str, CollectedData] = {}
+        stale_previous_records: list[CollectedData] = []
+
+        for existing_record in previous_current_records:
+            entity_key = existing_record.entity_key
+            if not entity_key:
+                continue
+            if entity_key not in previous_current_map:
+                previous_current_map[entity_key] = existing_record
+                continue
+            stale_previous_records.append(existing_record)
+
+        for stale_record in stale_previous_records:
+            stale_record.is_current = False
 
         for i, item in enumerate(data):
-            print(
-                f"DEBUG: Saving item {i}: name={item.get('name')}, metadata={item.get('metadata', 'NOT FOUND')}"
-            )
             raw_metadata = item.get("metadata", {})
             extra_data = build_dynamic_metadata(
                 raw_metadata,
@@ -318,20 +403,9 @@ class BaseCollector(ABC):
             previous_record = None
 
             if entity_key and entity_key not in seen_entity_keys:
-                result = await db.execute(
-                    select(CollectedData)
-                    .where(
-                        CollectedData.source == self.name,
-                        CollectedData.entity_key == entity_key,
-                        CollectedData.is_current == True,
-                    )
-                    .order_by(CollectedData.collected_at.desc().nullslast(), CollectedData.id.desc())
-                )
-                previous_records = result.scalars().all()
-                if previous_records:
-                    previous_record = previous_records[0]
-                    for old_record in previous_records:
-                        old_record.is_current = False
+                previous_record = previous_current_map.get(entity_key)
+                if previous_record is not None:
+                    previous_record.is_current = False
 
             record = CollectedData(
                 snapshot_id=snapshot_id,
@@ -375,7 +449,7 @@ class BaseCollector(ABC):
             seen_entity_keys.add(entity_key)
             records_added += 1
 
-            if i % 100 == 0:
+            if (i + 1) % progress_commit_interval == 0:
                 await self.update_progress(i + 1, commit=True)
 
         if snapshot_id is not None:

backend/app/services/collectors/celestrak.py

@@ -21,7 +21,7 @@ class CelesTrakTLECollector(BaseCollector):
 
     @property
     def base_url(self) -> str:
-        return "https://celestrak.org/NORAD/elements/gp.php"
+        return self._resolved_url or ""
 
     async def fetch(self) -> List[Dict[str, Any]]:
         satellite_groups = [
@@ -40,12 +40,15 @@ class CelesTrakTLECollector(BaseCollector):
         async with httpx.AsyncClient(timeout=120.0) as client:
             for group in satellite_groups:
                 try:
-                    url = f"https://celestrak.org/NORAD/elements/gp.php?GROUP={group}&FORMAT=json"
+                    url = f"{self.base_url}?GROUP={group}&FORMAT=json"
                     response = await client.get(url)
 
                     if response.status_code == 200:
                         data = response.json()
                         if isinstance(data, list):
+                            for item in data:
+                                if isinstance(item, dict):
+                                    item["_celestrak_group"] = group
                             all_satellites.extend(data)
                             print(f"CelesTrak: Fetched {len(data)} satellites from group '{group}'")
                 except Exception as e:
@@ -78,6 +81,7 @@ class CelesTrakTLECollector(BaseCollector):
                     "name": item.get("OBJECT_NAME", "Unknown"),
                     "reference_date": item.get("EPOCH", ""),
                     "metadata": {
+                        "constellation_group": item.get("_celestrak_group"),
                         "norad_cat_id": item.get("NORAD_CAT_ID"),
                         "international_designator": item.get("OBJECT_ID"),
                         "epoch": item.get("EPOCH"),

backend/app/services/collectors/cloudflare.py

@@ -39,6 +39,16 @@ class CloudflareRadarDeviceCollector(HTTPCollector):
         if CLOUDFLARE_API_TOKEN:
             self.headers["Authorization"] = f"Bearer {CLOUDFLARE_API_TOKEN}"
 
+    @property
+    def request_url(self) -> str:
+        return self._resolved_url or self.base_url
+
+    async def fetch(self) -> List[Dict[str, Any]]:
+        async with httpx.AsyncClient(timeout=60.0) as client:
+            response = await client.get(self.request_url, headers=self.headers)
+            response.raise_for_status()
+            return self.parse_response(response.json())
+
     def parse_response(self, response: Dict[str, Any]) -> List[Dict[str, Any]]:
         """Parse Cloudflare Radar device type response"""
         data = []
@@ -87,6 +97,16 @@ class CloudflareRadarTrafficCollector(HTTPCollector):
         if CLOUDFLARE_API_TOKEN:
             self.headers["Authorization"] = f"Bearer {CLOUDFLARE_API_TOKEN}"
 
+    @property
+    def request_url(self) -> str:
+        return self._resolved_url or self.base_url
+
+    async def fetch(self) -> List[Dict[str, Any]]:
+        async with httpx.AsyncClient(timeout=60.0) as client:
+            response = await client.get(self.request_url, headers=self.headers)
+            response.raise_for_status()
+            return self.parse_response(response.json())
+
     def parse_response(self, response: Dict[str, Any]) -> List[Dict[str, Any]]:
         """Parse Cloudflare Radar traffic timeseries response"""
         data = []
@@ -135,6 +155,16 @@ class CloudflareRadarTopASCollector(HTTPCollector):
         if CLOUDFLARE_API_TOKEN:
             self.headers["Authorization"] = f"Bearer {CLOUDFLARE_API_TOKEN}"
 
+    @property
+    def request_url(self) -> str:
+        return self._resolved_url or self.base_url
+
+    async def fetch(self) -> List[Dict[str, Any]]:
+        async with httpx.AsyncClient(timeout=60.0) as client:
+            response = await client.get(self.request_url, headers=self.headers)
+            response.raise_for_status()
+            return self.parse_response(response.json())
+
     def parse_response(self, response: Dict[str, Any]) -> List[Dict[str, Any]]:
         """Parse Cloudflare Radar top locations response"""
         data = []

backend/app/services/collectors/epoch_ai.py

@@ -23,7 +23,7 @@ class EpochAIGPUCollector(BaseCollector):
 
     async def fetch(self) -> List[Dict[str, Any]]:
         """Fetch Epoch AI GPU clusters data from webpage"""
-        url = "https://epoch.ai/data/gpu-clusters"
+        url = self._resolved_url or ""
 
         async with httpx.AsyncClient(timeout=60.0) as client:
             response = await client.get(url)

backend/app/services/collectors/fao_landing.py

@@ -18,11 +18,9 @@ class FAOLandingPointCollector(BaseCollector):
     frequency_hours = 168
     data_type = "landing_point"
 
-    csv_url = "https://data.apps.fao.org/catalog/dataset/1b75ff21-92f2-4b96-9b7b-98e8aa65ad5d/resource/b6071077-d1d4-4e97-aa00-42e902847c87/download/landing-point-geo.csv"
-
     async def fetch(self) -> List[Dict[str, Any]]:
         async with httpx.AsyncClient(timeout=60.0) as client:
-            response = await client.get(self.csv_url)
+            response = await client.get(self._resolved_url or "")
             response.raise_for_status()
             return self.parse_csv(response.text)
 

backend/app/services/collectors/huggingface.py

@@ -21,6 +21,18 @@ class HuggingFaceModelCollector(HTTPCollector):
     data_type = "model"
     base_url = "https://huggingface.co/api/models"
 
+    @property
+    def request_url(self) -> str:
+        return self._resolved_url or self.base_url
+
+    async def fetch(self) -> List[Dict[str, Any]]:
+        from httpx import AsyncClient
+
+        async with AsyncClient(timeout=60.0) as client:
+            response = await client.get(self.request_url, headers=self.headers)
+            response.raise_for_status()
+            return self.parse_response(response.json())
+
     def parse_response(self, response: Dict[str, Any]) -> List[Dict[str, Any]]:
         """Parse Hugging Face models API response"""
         data = []
@@ -63,6 +75,18 @@ class HuggingFaceDatasetCollector(HTTPCollector):
     data_type = "dataset"
     base_url = "https://huggingface.co/api/datasets"
 
+    @property
+    def request_url(self) -> str:
+        return self._resolved_url or self.base_url
+
+    async def fetch(self) -> List[Dict[str, Any]]:
+        from httpx import AsyncClient
+
+        async with AsyncClient(timeout=60.0) as client:
+            response = await client.get(self.request_url, headers=self.headers)
+            response.raise_for_status()
+            return self.parse_response(response.json())
+
     def parse_response(self, response: Dict[str, Any]) -> List[Dict[str, Any]]:
         """Parse Hugging Face datasets API response"""
         data = []
@@ -104,6 +128,18 @@ class HuggingFaceSpacesCollector(HTTPCollector):
     data_type = "space"
     base_url = "https://huggingface.co/api/spaces"
 
+    @property
+    def request_url(self) -> str:
+        return self._resolved_url or self.base_url
+
+    async def fetch(self) -> List[Dict[str, Any]]:
+        from httpx import AsyncClient
+
+        async with AsyncClient(timeout=60.0) as client:
+            response = await client.get(self.request_url, headers=self.headers)
+            response.raise_for_status()
+            return self.parse_response(response.json())
+
     def parse_response(self, response: Dict[str, Any]) -> List[Dict[str, Any]]:
         """Parse Hugging Face Spaces API response"""
         data = []

backend/app/services/collectors/news_live_streams.py

@@ -0,0 +1,570 @@
+from __future__ import annotations
+
+import asyncio
+import base64
+from datetime import UTC, datetime
+from typing import Any
+from urllib.parse import urlparse
+
+import httpx
+from sqlalchemy import select
+
+from app.core.data_sources import get_data_sources_config
+from app.models.datasource_config import DataSourceConfig
+from app.services.collectors.base import BaseCollector
+
+
+class NewsLiveStreamsCollector(BaseCollector):
+    """Collect normalized news live-stream sources from a JSON endpoint."""
+
+    name = "news_live_streams"
+    priority = "P2"
+    module = "L4"
+    frequency_hours = 12
+    data_type = "news_live_stream"
+    fail_on_empty = False
+
+    DEFAULT_TIMEOUT = 45.0
+    DEFAULT_HEADERS = {
+        "User-Agent": "Planet-Intelligence-System/1.0 (Python/collector)",
+        "Accept": "application/json",
+    }
+    RESPONSE_CANDIDATE_KEYS = ("sources", "streams", "channels", "items", "results", "data")
+    DEFAULT_ADAPTER = "iptv_org"
+    DEFAULT_IPTV_ORG_STREAMS_URL = "https://iptv-org.github.io/api/streams.json"
+    DEFAULT_IPTV_ORG_LOGOS_URL = "https://iptv-org.github.io/api/logos.json"
+    DEFAULT_IPTV_ORG_NEWS_CATEGORIES = ("news", "business", "weather")
+    DEFAULT_IPTV_ORG_EXCLUDE_CATEGORIES = ("music", "sports", "kids", "entertainment")
+    DEFAULT_IPTV_ORG_MAX_SOURCES = 120
+
+    async def fetch(self) -> list[dict[str, Any]]:
+        request_url = (self._resolved_url or "").strip()
+        if not request_url:
+            return []
+
+        datasource_config = await self._load_datasource_config()
+        effective_config = self._get_effective_config(datasource_config)
+        adapter = str(effective_config.get("adapter") or "").strip().lower()
+        if adapter == "iptv_org":
+            return await self._fetch_iptv_org(request_url, effective_config)
+
+        request_headers = self._build_request_headers(datasource_config)
+        request_config = self._get_request_config(datasource_config)
+        request_params = self._build_request_params(datasource_config)
+        request_json = self._build_request_json_body(datasource_config)
+        request_data = self._build_request_form_body(datasource_config)
+        timeout = self._get_timeout(datasource_config)
+
+        async with httpx.AsyncClient(timeout=timeout, follow_redirects=True) as client:
+            response = await client.request(
+                request_config["method"],
+                request_url,
+                headers=request_headers,
+                params=request_params or None,
+                json=request_json,
+                data=request_data,
+            )
+            response.raise_for_status()
+            return self.parse_response(
+                response.json(),
+                response_path=request_config["response_path"],
+            )
+
+    async def _load_datasource_config(self) -> DataSourceConfig | None:
+        if not self._db_session:
+            return None
+
+        result = await self._db_session.execute(
+            select(DataSourceConfig)
+            .where(DataSourceConfig.name == self.name)
+            .where(DataSourceConfig.is_active.is_(True))
+            .limit(1)
+        )
+        return result.scalar_one_or_none()
+
+    def _get_effective_config(self, datasource_config: DataSourceConfig | None) -> dict[str, Any]:
+        payload = dict(datasource_config.config or {}) if datasource_config else {}
+        if payload:
+            return payload
+
+        yaml_config = get_data_sources_config()
+        return {
+            "adapter": self.DEFAULT_ADAPTER,
+            "streams_url": yaml_config.get_yaml_value("news_live_streams.streams_url")
+            or self.DEFAULT_IPTV_ORG_STREAMS_URL,
+            "logos_url": yaml_config.get_yaml_value("news_live_streams.logos_url")
+            or self.DEFAULT_IPTV_ORG_LOGOS_URL,
+            "news_categories": list(self.DEFAULT_IPTV_ORG_NEWS_CATEGORIES),
+            "exclude_categories": list(self.DEFAULT_IPTV_ORG_EXCLUDE_CATEGORIES),
+            "max_sources": self.DEFAULT_IPTV_ORG_MAX_SOURCES,
+        }
+
+    def _get_request_config(self, datasource_config: DataSourceConfig | None) -> dict[str, Any]:
+        payload = self._get_effective_config(datasource_config)
+        raw_method = payload.get("method") or payload.get("request_method") or "GET"
+        method = str(raw_method).strip().upper() or "GET"
+        if method not in {"GET", "POST"}:
+            method = "GET"
+
+        response_path = payload.get("response_path") or payload.get("payload_path") or payload.get("items_path")
+        if isinstance(response_path, str):
+            response_path = response_path.strip()
+        else:
+            response_path = None
+
+        return {
+            "method": method,
+            "response_path": response_path or None,
+        }
+
+    def _get_timeout(self, datasource_config: DataSourceConfig | None) -> float:
+        payload = self._get_effective_config(datasource_config)
+        try:
+            return float(payload.get("timeout", self.DEFAULT_TIMEOUT))
+        except (TypeError, ValueError):
+            return self.DEFAULT_TIMEOUT
+
+    def _build_request_headers(self, datasource_config: DataSourceConfig | None) -> dict[str, str]:
+        headers = dict(self.DEFAULT_HEADERS)
+        if datasource_config:
+            headers.update(self._normalize_headers(datasource_config.headers))
+            headers.update(self._build_auth_headers(datasource_config))
+        return headers
+
+    def _build_request_params(self, datasource_config: DataSourceConfig | None) -> dict[str, Any]:
+        params: dict[str, Any] = {}
+        if not datasource_config:
+            return params
+
+        payload = datasource_config.config or {}
+        candidate = payload.get("params") or payload.get("query_params")
+        if isinstance(candidate, dict):
+            params.update(candidate)
+
+        if datasource_config.auth_type == "api_key":
+            auth_config = datasource_config.auth_config or {}
+            if str(auth_config.get("in") or auth_config.get("location") or "header").lower() == "query":
+                api_key = auth_config.get("api_key")
+                key_name = auth_config.get("key_name") or auth_config.get("param_name") or "api_key"
+                if api_key and key_name:
+                    params[str(key_name)] = api_key
+
+        return params
+
+    def _build_request_json_body(self, datasource_config: DataSourceConfig | None) -> Any:
+        if not datasource_config:
+            return None
+
+        payload = datasource_config.config or {}
+        body = payload.get("json_body")
+        if body is None and str(payload.get("body_type") or "").lower() in {"json", ""}:
+            candidate = payload.get("body")
+            if isinstance(candidate, (dict, list)):
+                body = candidate
+        return body
+
+    def _build_request_form_body(self, datasource_config: DataSourceConfig | None) -> Any:
+        if not datasource_config:
+            return None
+
+        payload = datasource_config.config or {}
+        form_body = payload.get("form_body")
+        if form_body is not None:
+            return form_body
+
+        if str(payload.get("body_type") or "").lower() == "form":
+            candidate = payload.get("body")
+            if isinstance(candidate, dict):
+                return candidate
+        return None
+
+    def _normalize_headers(self, headers: Any) -> dict[str, str]:
+        if not isinstance(headers, dict):
+            return {}
+        normalized: dict[str, str] = {}
+        for key, value in headers.items():
+            header_name = str(key).strip()
+            if not header_name or value is None:
+                continue
+            normalized[header_name] = str(value)
+        return normalized
+
+    def _build_auth_headers(self, datasource_config: DataSourceConfig | None) -> dict[str, str]:
+        if not datasource_config:
+            return {}
+
+        auth_type = str(datasource_config.auth_type or "none").lower()
+        auth_config = datasource_config.auth_config or {}
+        if auth_type == "bearer" and auth_config.get("token"):
+            return {"Authorization": f"Bearer {auth_config['token']}"}
+
+        if auth_type == "api_key" and auth_config.get("api_key"):
+            location = str(auth_config.get("in") or auth_config.get("location") or "header").lower()
+            if location == "query":
+                return {}
+            key_name = auth_config.get("key_name") or "X-API-Key"
+            return {str(key_name): str(auth_config["api_key"])}
+
+        if auth_type == "basic":
+            username = str(auth_config.get("username") or "")
+            password = str(auth_config.get("password") or "")
+            encoded = base64.b64encode(f"{username}:{password}".encode()).decode()
+            return {"Authorization": f"Basic {encoded}"}
+
+        return {}
+
+    def _extract_candidates(self, response: Any, response_path: str | None) -> list[Any]:
+        if response_path:
+            extracted = self._extract_from_path(response, response_path)
+            if isinstance(extracted, list):
+                return extracted
+            if isinstance(extracted, dict):
+                for key in self.RESPONSE_CANDIDATE_KEYS:
+                    nested = extracted.get(key)
+                    if isinstance(nested, list):
+                        return nested
+                return [extracted]
+
+        if isinstance(response, dict):
+            for key in self.RESPONSE_CANDIDATE_KEYS:
+                nested = response.get(key)
+                if isinstance(nested, list):
+                    return nested
+            return []
+
+        if isinstance(response, list):
+            return response
+        return []
+
+    def _extract_from_path(self, payload: Any, path: str) -> Any:
+        current = payload
+        for segment in (part.strip() for part in path.split(".") if part.strip()):
+            if isinstance(current, dict):
+                current = current.get(segment)
+                continue
+            if isinstance(current, list):
+                try:
+                    current = current[int(segment)]
+                except (TypeError, ValueError, IndexError):
+                    return None
+                continue
+            return None
+        return current
+
+    def _infer_source_type(self, item: dict[str, Any]) -> str:
+        explicit = str(item.get("source_type") or item.get("type") or "").strip().lower()
+        if explicit in {"iframe", "hls", "video", "external", "youtube"}:
+            return explicit
+
+        youtube_video_id = self._clean_text(
+            item.get("youtube_video_id")
+            or item.get("video_id")
+            or item.get("youtubeVideoId")
+        )
+        youtube_channel = self._clean_text(item.get("youtube_channel") or item.get("channel_handle"))
+        embed_url = self._clean_url(item.get("embed_url") or item.get("embed") or item.get("page_url"))
+        stream_url = self._clean_url(item.get("stream_url") or item.get("stream") or item.get("playback_url") or item.get("hls_url"))
+        homepage_url = self._clean_url(item.get("homepage_url") or item.get("source_url") or item.get("website"))
+
+        if youtube_video_id or youtube_channel:
+            return "youtube"
+        if stream_url.endswith(".m3u8"):
+            return "hls"
+        if stream_url:
+            return "video"
+        if embed_url:
+            parsed = urlparse(embed_url)
+            if "youtube.com" in (parsed.netloc or "") or "youtu.be" in (parsed.netloc or ""):
+                return "youtube"
+            return "iframe"
+        if homepage_url:
+            return "external"
+        return "iframe"
+
+    def _parse_enabled(self, item: dict[str, Any]) -> bool:
+        if "is_enabled" in item:
+            return self._to_bool(item.get("is_enabled"), default=True)
+        if "enabled" in item:
+            return self._to_bool(item.get("enabled"), default=True)
+        if "active" in item:
+            return self._to_bool(item.get("active"), default=True)
+        if "status" in item:
+            status = str(item.get("status") or "").strip().lower()
+            if status in {"disabled", "inactive", "offline"}:
+                return False
+            if status in {"enabled", "active", "online", "live"}:
+                return True
+        return True
+
+    def _to_bool(self, value: Any, *, default: bool) -> bool:
+        if isinstance(value, bool):
+            return value
+        if value in (None, ""):
+            return default
+        if isinstance(value, str):
+            lowered = value.strip().lower()
+            if lowered in {"1", "true", "yes", "on", "enabled", "active", "online", "live"}:
+                return True
+            if lowered in {"0", "false", "no", "off", "disabled", "inactive", "offline"}:
+                return False
+        return bool(value)
+
+    def _clean_text(self, value: Any) -> str:
+        if value is None:
+            return ""
+        return str(value).strip()
+
+    def _clean_url(self, value: Any) -> str:
+        text = self._clean_text(value)
+        if not text:
+            return ""
+        parsed = urlparse(text)
+        if parsed.scheme and parsed.scheme not in {"http", "https"}:
+            return ""
+        if parsed.scheme and not parsed.netloc:
+            return ""
+        return text
+
+    async def _fetch_iptv_org(self, channels_url: str, collector_config: dict[str, Any]) -> list[dict[str, Any]]:
+        streams_url = self._clean_url(collector_config.get("streams_url")) or self.DEFAULT_IPTV_ORG_STREAMS_URL
+        logos_url = self._clean_url(collector_config.get("logos_url")) or self.DEFAULT_IPTV_ORG_LOGOS_URL
+        news_categories = {
+            self._clean_text(value).lower()
+            for value in (collector_config.get("news_categories") or self.DEFAULT_IPTV_ORG_NEWS_CATEGORIES)
+            if self._clean_text(value)
+        }
+        exclude_categories = {
+            self._clean_text(value).lower()
+            for value in (collector_config.get("exclude_categories") or self.DEFAULT_IPTV_ORG_EXCLUDE_CATEGORIES)
+            if self._clean_text(value)
+        }
+        try:
+            max_sources = int(collector_config.get("max_sources", self.DEFAULT_IPTV_ORG_MAX_SOURCES))
+        except (TypeError, ValueError):
+            max_sources = self.DEFAULT_IPTV_ORG_MAX_SOURCES
+
+        timeout = self.DEFAULT_TIMEOUT
+        try:
+            timeout = float(collector_config.get("timeout", self.DEFAULT_TIMEOUT))
+        except (TypeError, ValueError):
+            timeout = self.DEFAULT_TIMEOUT
+
+        async with httpx.AsyncClient(timeout=timeout, follow_redirects=True) as client:
+            channels_payload, streams_payload, logos_payload = await self._gather_iptv_org_payloads(
+                client,
+                channels_url,
+                streams_url,
+                logos_url,
+            )
+
+        channels = channels_payload if isinstance(channels_payload, list) else []
+        streams = streams_payload if isinstance(streams_payload, list) else []
+        logos = logos_payload if isinstance(logos_payload, list) else []
+
+        logo_by_channel = {
+            self._clean_text(item.get("channel")): self._clean_url(item.get("url"))
+            for item in logos
+            if isinstance(item, dict) and self._clean_text(item.get("channel")) and self._clean_url(item.get("url"))
+        }
+
+        streams_by_channel: dict[str, list[dict[str, Any]]] = {}
+        for stream in streams:
+            if not isinstance(stream, dict):
+                continue
+            channel_id = self._clean_text(stream.get("channel"))
+            if not channel_id:
+                continue
+            streams_by_channel.setdefault(channel_id, []).append(stream)
+
+        normalized: list[dict[str, Any]] = []
+        for channel in channels:
+            if not isinstance(channel, dict):
+                continue
+
+            categories = [
+                self._clean_text(value).lower()
+                for value in (channel.get("categories") or [])
+                if self._clean_text(value)
+            ]
+            if news_categories and not any(category in news_categories for category in categories):
+                continue
+            if exclude_categories and any(category in exclude_categories for category in categories):
+                continue
+            if channel.get("is_nsfw") is True:
+                continue
+            if channel.get("closed"):
+                continue
+
+            channel_id = self._clean_text(channel.get("id"))
+            if not channel_id:
+                continue
+
+            stream = self._pick_iptv_org_stream(streams_by_channel.get(channel_id) or [])
+            if not stream:
+                continue
+
+            stream_url = self._clean_url(stream.get("url"))
+            if not stream_url:
+                continue
+
+            name = self._clean_text(channel.get("name")) or channel_id
+            notes_parts = [
+                f"Imported from IPTV-org catalog ({channel_id})",
+                f"Categories: {', '.join(categories)}" if categories else "",
+                f"Quality: {self._clean_text(stream.get('quality'))}" if self._clean_text(stream.get("quality")) else "",
+            ]
+            metadata = {
+                "provider": self._clean_text(channel.get("network")) or "IPTV-org",
+                "region": self._clean_text(channel.get("country")) or "Global",
+                "language": "und",
+                "source_type": "hls" if stream_url.endswith(".m3u8") else "video",
+                "embed_url": "",
+                "stream_url": stream_url,
+                "homepage_url": self._clean_url(channel.get("website")),
+                "poster_url": logo_by_channel.get(channel_id, ""),
+                "youtube_video_id": "",
+                "youtube_channel": "",
+                "sort_order": 400 + len(normalized),
+                "notes": "; ".join(part for part in notes_parts if part),
+                "is_enabled": True,
+                "collector_adapter": "iptv_org",
+                "channel_id": channel_id,
+                "categories": categories,
+                "quality": self._clean_text(stream.get("quality")),
+                "stream_label": self._clean_text(stream.get("label") or stream.get("title")),
+                "stream_referrer": self._clean_text(stream.get("referrer")),
+                "stream_user_agent": self._clean_text(stream.get("user_agent")),
+            }
+
+            normalized.append(
+                {
+                    "source_id": channel_id,
+                    "name": name,
+                    "description": metadata["notes"],
+                    "metadata": metadata,
+                    "reference_date": datetime.now(UTC).isoformat(),
+                }
+            )
+            if len(normalized) >= max_sources:
+                break
+
+        return normalized
+
+    async def _gather_iptv_org_payloads(
+        self,
+        client: httpx.AsyncClient,
+        channels_url: str,
+        streams_url: str,
+        logos_url: str,
+    ) -> tuple[Any, Any, Any]:
+        headers = dict(self.DEFAULT_HEADERS)
+        channels_payload, streams_payload, logos_payload = await asyncio.gather(
+            client.get(channels_url, headers=headers),
+            client.get(streams_url, headers=headers),
+            client.get(logos_url, headers=headers),
+        )
+        channels_payload.raise_for_status()
+        streams_payload.raise_for_status()
+        logos_payload.raise_for_status()
+        return channels_payload.json(), streams_payload.json(), logos_payload.json()
+
+    def _pick_iptv_org_stream(self, streams: list[dict[str, Any]]) -> dict[str, Any] | None:
+        if not streams:
+            return None
+
+        def score(stream: dict[str, Any]) -> tuple[int, int]:
+            url = self._clean_url(stream.get("url"))
+            quality = self._clean_text(stream.get("quality")).lower()
+            quality_score = 0
+            if quality.endswith("p"):
+                try:
+                    quality_score = int(quality[:-1])
+                except ValueError:
+                    quality_score = 0
+            stream_score = 1000 if url.endswith(".m3u8") else 0
+            return stream_score, quality_score
+
+        sorted_streams = sorted(streams, key=score, reverse=True)
+        return sorted_streams[0]
+
+    def parse_response(self, response: Any, *, response_path: str | None = None) -> list[dict[str, Any]]:
+        candidates = self._extract_candidates(response, response_path)
+
+        normalized: list[dict[str, Any]] = []
+        for index, item in enumerate(candidates):
+            if not isinstance(item, dict):
+                continue
+
+            stream_id = (
+                item.get("id")
+                or item.get("source_id")
+                or item.get("slug")
+                or item.get("channel_id")
+                or item.get("code")
+                or f"news-live-{index + 1}"
+            )
+            name = self._clean_text(
+                item.get("name")
+                or item.get("title")
+                or item.get("channel")
+                or item.get("display_name")
+                or f"News Live {index + 1}"
+            )
+            if not name:
+                continue
+
+            source_type = self._infer_source_type(item)
+            stream_url = self._clean_url(
+                item.get("stream_url")
+                or item.get("stream")
+                or item.get("playback_url")
+                or item.get("hls_url")
+                or item.get("m3u8_url")
+            )
+            embed_url = self._clean_url(
+                item.get("embed_url")
+                or item.get("embed")
+                or item.get("page_url")
+                or (item.get("url") if source_type == "iframe" else "")
+            )
+            homepage_url = self._clean_url(
+                item.get("homepage_url")
+                or item.get("source_url")
+                or item.get("website")
+                or item.get("url")
+            )
+            metadata = {
+                "provider": self._clean_text(item.get("provider") or item.get("publisher") or item.get("network")) or "Collector",
+                "region": self._clean_text(item.get("region") or item.get("country") or item.get("market")) or "Global",
+                "language": self._clean_text(item.get("language") or item.get("lang") or item.get("locale")) or "und",
+                "source_type": source_type,
+                "embed_url": embed_url,
+                "stream_url": stream_url,
+                "homepage_url": homepage_url,
+                "poster_url": self._clean_url(item.get("poster_url") or item.get("thumbnail_url") or item.get("logo_url")),
+                "youtube_video_id": self._clean_text(
+                    item.get("youtube_video_id")
+                    or item.get("video_id")
+                    or item.get("youtubeVideoId")
+                ),
+                "youtube_channel": self._clean_text(
+                    item.get("youtube_channel")
+                    or item.get("channel_handle")
+                    or item.get("youtubeChannel")
+                ),
+                "sort_order": item.get("sort_order", 200 + index),
+                "notes": self._clean_text(item.get("notes") or item.get("description") or item.get("summary")),
+                "is_enabled": self._parse_enabled(item),
+            }
+
+            normalized.append(
+                {
+                    "source_id": str(stream_id),
+                    "name": name,
+                    "description": metadata["notes"],
+                    "metadata": metadata,
+                    "reference_date": item.get("reference_date") or datetime.now(UTC).isoformat(),
+                }
+            )
+
+        return normalized

backend/app/services/collectors/peeringdb.py

@@ -16,6 +16,7 @@ from typing import Dict, Any, List
 from datetime import UTC, datetime
 
 import httpx
+from urllib.parse import urlencode
 from app.services.collectors.base import HTTPCollector
 
 
@@ -38,9 +39,13 @@ class PeeringDBIXPCollector(HTTPCollector):
             "User-Agent": "Planet-Intelligence-System/1.0 (Python/collector)",
             "Accept": "application/json",
         }
-        # API key is added to URL as query parameter
-        if PEERINGDB_API_KEY:
-            self.base_url = f"{self.base_url}?key={PEERINGDB_API_KEY}"
+    @property
+    def request_url(self) -> str:
+        base = self._resolved_url or self.base_url
+        if not PEERINGDB_API_KEY:
+            return base
+        separator = "&" if "?" in base else "?"
+        return f"{base}{separator}{urlencode({'key': PEERINGDB_API_KEY})}"
 
     async def fetch_with_retry(
         self, max_retries: int = 3, base_delay: float = 2.0
@@ -51,7 +56,7 @@ class PeeringDBIXPCollector(HTTPCollector):
         for attempt in range(max_retries):
             try:
                 async with httpx.AsyncClient(timeout=60.0) as client:
-                    response = await client.get(self.base_url, headers=self.headers)
+                    response = await client.get(self.request_url, headers=self.headers)
 
                     if response.status_code == 429:
                         # Rate limited - wait and retry with exponential backoff
@@ -141,8 +146,13 @@ class PeeringDBNetworkCollector(HTTPCollector):
             "User-Agent": "Planet-Intelligence-System/1.0 (Python/collector)",
             "Accept": "application/json",
         }
-        if PEERINGDB_API_KEY:
-            self.base_url = f"{self.base_url}?key={PEERINGDB_API_KEY}"
+    @property
+    def request_url(self) -> str:
+        base = self._resolved_url or self.base_url
+        if not PEERINGDB_API_KEY:
+            return base
+        separator = "&" if "?" in base else "?"
+        return f"{base}{separator}{urlencode({'key': PEERINGDB_API_KEY})}"
 
     async def fetch_with_retry(
         self, max_retries: int = 3, base_delay: float = 2.0
@@ -153,7 +163,7 @@ class PeeringDBNetworkCollector(HTTPCollector):
         for attempt in range(max_retries):
             try:
                 async with httpx.AsyncClient(timeout=60.0) as client:
-                    response = await client.get(self.base_url, headers=self.headers)
+                    response = await client.get(self.request_url, headers=self.headers)
 
                     if response.status_code == 429:
                         delay = base_delay * (2**attempt)
@@ -244,8 +254,13 @@ class PeeringDBFacilityCollector(HTTPCollector):
             "User-Agent": "Planet-Intelligence-System/1.0 (Python/collector)",
             "Accept": "application/json",
         }
-        if PEERINGDB_API_KEY:
-            self.base_url = f"{self.base_url}?key={PEERINGDB_API_KEY}"
+    @property
+    def request_url(self) -> str:
+        base = self._resolved_url or self.base_url
+        if not PEERINGDB_API_KEY:
+            return base
+        separator = "&" if "?" in base else "?"
+        return f"{base}{separator}{urlencode({'key': PEERINGDB_API_KEY})}"
 
     async def fetch_with_retry(
         self, max_retries: int = 3, base_delay: float = 2.0
@@ -256,7 +271,7 @@ class PeeringDBFacilityCollector(HTTPCollector):
         for attempt in range(max_retries):
             try:
                 async with httpx.AsyncClient(timeout=60.0) as client:
-                    response = await client.get(self.base_url, headers=self.headers)
+                    response = await client.get(self.request_url, headers=self.headers)
 
                     if response.status_code == 429:
                         delay = base_delay * (2**attempt)

backend/app/services/collectors/ris_live.py

@@ -33,7 +33,7 @@ class RISLiveCollector(BaseCollector):
 
     def _fetch_via_stream(self) -> list[dict[str, Any]]:
         events: list[dict[str, Any]] = []
-        stream_url = "https://ris-live.ripe.net/v1/stream/?format=json&client=planet-ris-live"
+        stream_url = self._resolved_url or ""
         subscribe = json.dumps(
             {
                 "host": "rrc00",

backend/app/services/collectors/spacetrack.py

@@ -7,6 +7,7 @@ API documentation: https://www.space-track.org/documentation
 import json
 from typing import Dict, Any, List
 import httpx
+from urllib.parse import urlparse
 
 from app.services.collectors.base import BaseCollector
 from app.core.data_sources import get_data_sources_config
@@ -21,12 +22,30 @@ class SpaceTrackTLECollector(BaseCollector):
     data_type = "satellite_tle"
 
     @property
-    def base_url(self) -> str:
+    def query_url(self) -> str:
         config = get_data_sources_config()
         if self._resolved_url:
             return self._resolved_url
         return config.get_yaml_url("spacetrack_tle")
 
+    @property
+    def site_root(self) -> str:
+        config = get_data_sources_config()
+        configured_root = config.get_yaml_value("spacetrack.base_url")
+        if isinstance(configured_root, str) and configured_root:
+            return configured_root.rstrip("/")
+
+        parsed = urlparse(self.query_url)
+        return f"{parsed.scheme}://{parsed.netloc}".rstrip("/")
+
+    @property
+    def login_url(self) -> str:
+        return f"{self.site_root}/ajaxauth/login"
+
+    @property
+    def probe_url(self) -> str:
+        return f"{self.site_root}/basicspacedata/query/class/gp/NORAD_CAT_ID/25544/format/json"
+
     async def fetch(self) -> List[Dict[str, Any]]:
         from app.core.config import settings
 
@@ -47,13 +66,13 @@ class SpaceTrackTLECollector(BaseCollector):
                     "User-Agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/120.0.0.0 Safari/537.36",
                     "Accept": "application/json, text/html, */*",
                     "Accept-Language": "en-US,en;q=0.9",
-                    "Referer": "https://www.space-track.org/",
+                    "Referer": f"{self.site_root}/",
                 },
             ) as client:
-                await client.get("https://www.space-track.org/")
+                await client.get(f"{self.site_root}/")
 
                 login_response = await client.post(
-                    "https://www.space-track.org/ajaxauth/login",
+                    self.login_url,
                     data={
                         "identity": username,
                         "password": password,
@@ -69,7 +88,7 @@ class SpaceTrackTLECollector(BaseCollector):
                         timeout=120.0,
                         follow_redirects=True,
                     ) as alt_client:
-                        await alt_client.get("https://www.space-track.org/")
+                        await alt_client.get(f"{self.site_root}/")
 
                         form_data = {
                             "username": username,
@@ -77,7 +96,7 @@ class SpaceTrackTLECollector(BaseCollector):
                             "query": "class/gp/NORAD_CAT_ID/25544/format/json",
                         }
                         alt_login = await alt_client.post(
-                            "https://www.space-track.org/ajaxauth/login",
+                            self.login_url,
                             data={
                                 "identity": username,
                                 "password": password,
@@ -86,9 +105,7 @@ class SpaceTrackTLECollector(BaseCollector):
                         print(f"SPACETRACK: Alt login status: {alt_login.status_code}")
 
                         if alt_login.status_code == 200:
-                            tle_response = await alt_client.get(
-                                "https://www.space-track.org/basicspacedata/query/class/gp/NORAD_CAT_ID/25544/format/json"
-                            )
+                            tle_response = await alt_client.get(self.probe_url)
                             if tle_response.status_code == 200:
                                 data = tle_response.json()
                                 print(f"SPACETRACK: Received {len(data)} records via alt method")
@@ -98,9 +115,7 @@ class SpaceTrackTLECollector(BaseCollector):
                     print(f"SPACETRACK: Login failed, using sample data")
                     return self._get_sample_data()
 
-                tle_response = await client.get(
-                    "https://www.space-track.org/basicspacedata/query/class/gp/NORAD_CAT_ID/25544/format/json"
-                )
+                tle_response = await client.get(self.probe_url)
                 print(f"SPACETRACK: TLE query status: {tle_response.status_code}")
 
                 if tle_response.status_code != 200:
@@ -127,11 +142,11 @@ class SpaceTrackTLECollector(BaseCollector):
                 },
             ) as client:
                 # First, visit the main page to get any cookies
-                await client.get("https://www.space-track.org/")
+                await client.get(f"{self.site_root}/")
 
                 # Login to get session cookie
                 login_response = await client.post(
-                    "https://www.space-track.org/ajaxauth/login",
+                    self.login_url,
                     data={
                         "identity": username,
                         "password": password,
@@ -146,13 +161,7 @@ class SpaceTrackTLECollector(BaseCollector):
                     return self._get_sample_data()
 
                 # Query for TLE data (get first 1000 satellites)
-                tle_response = await client.get(
-                    "https://www.space-track.org/basicspacedata/query"
-                    "/class/gp"
-                    "/orderby/EPOCH%20desc"
-                    "/limit/1000"
-                    "/format/json"
-                )
+                tle_response = await client.get(self.query_url)
                 print(f"SPACETRACK: TLE query status: {tle_response.status_code}")
 
                 if tle_response.status_code != 200:

backend/app/services/collectors/telegeography.py

@@ -11,6 +11,7 @@ from datetime import UTC, datetime
 from bs4 import BeautifulSoup
 import httpx
 
+from app.core.data_sources import get_data_sources_config
 from app.services.collectors.base import BaseCollector
 
 
@@ -24,15 +25,17 @@ class TeleGeographyCableCollector(BaseCollector):
 
     async def fetch(self) -> List[Dict[str, Any]]:
         """Fetch submarine cable data from Wayback Machine"""
+        config = get_data_sources_config()
         # Try multiple data sources
         sources = [
-            # Wayback Machine archive of TeleGeography
-            "https://web.archive.org/web/2024/https://www.submarinecablemap.com/api/v3/cable",
-            # Alternative: Try scraping the page
-            "https://www.submarinecablemap.com",
+            self._resolved_url or "",
+            str(config.get_yaml_value("telegeography.archived_cable_url") or ""),
+            str(config.get_yaml_value("telegeography.live_map_url") or ""),
         ]
 
         for url in sources:
+            if not url:
+                continue
             try:
                 async with httpx.AsyncClient(timeout=60.0, follow_redirects=True) as client:
                     response = await client.get(url)
@@ -161,7 +164,7 @@ class TeleGeographyLandingPointCollector(BaseCollector):
 
     async def fetch(self) -> List[Dict[str, Any]]:
         """Fetch landing point data from GitHub mirror"""
-        url = "https://raw.githubusercontent.com/lintaojlu/submarine_cable_information/main/landing_point.json"
+        url = self._resolved_url or ""
 
         async with httpx.AsyncClient(timeout=60.0) as client:
             response = await client.get(url)
@@ -225,7 +228,7 @@ class TeleGeographyCableSystemCollector(BaseCollector):
 
     async def fetch(self) -> List[Dict[str, Any]]:
         """Fetch cable system data"""
-        url = "https://raw.githubusercontent.com/lintaojlu/submarine_cable_information/main/cable.json"
+        url = self._resolved_url or ""
 
         async with httpx.AsyncClient(timeout=60.0) as client:
             response = await client.get(url)

backend/app/services/collectors/top500.py

@@ -10,6 +10,7 @@ from typing import Dict, Any, List
 from bs4 import BeautifulSoup
 import httpx
 
+from app.core.data_sources import get_data_sources_config
 from app.services.collectors.base import BaseCollector
 
 
@@ -22,7 +23,7 @@ class TOP500Collector(BaseCollector):
 
     async def fetch(self) -> List[Dict[str, Any]]:
         """Fetch TOP500 list data and enrich each row with detail-page metadata."""
-        url = "https://top500.org/lists/top500/list/2025/11/"
+        url = self._resolved_url or ""
 
         async with httpx.AsyncClient(timeout=60.0, follow_redirects=True) as client:
             response = await client.get(url)
@@ -48,11 +49,13 @@ class TOP500Collector(BaseCollector):
             return await asyncio.gather(*(enrich(entry) for entry in entries))
 
     def _extract_system_fields(self, system_cell) -> Dict[str, str]:
+        config = get_data_sources_config()
+        top500_base_url = config.get_yaml_value("top500.base_url") or "https://top500.org"
         link = system_cell.find("a")
         system_name = link.get_text(" ", strip=True) if link else system_cell.get_text(" ", strip=True)
         detail_url = ""
         if link and link.get("href"):
-            detail_url = f"https://top500.org{link.get('href')}"
+            detail_url = f"{str(top500_base_url).rstrip('/')}{link.get('href')}"
 
         manufacturer = ""
         if link and link.next_sibling:

backend/app/services/earth_news.py

@@ -0,0 +1,540 @@
+from __future__ import annotations
+
+import asyncio
+from dataclasses import dataclass
+from datetime import UTC, datetime
+from email.utils import parsedate_to_datetime
+import hashlib
+import html
+import re
+from typing import Any
+from urllib.parse import quote
+import xml.etree.ElementTree as ET
+
+import httpx
+from bs4 import BeautifulSoup
+
+
+USER_AGENT = "PlanetEarthNewsBoard/1.0 (+https://planet.local)"
+REQUEST_TIMEOUT = 12.0
+MAX_ITEMS_PER_SOURCE = 6
+MAX_ITEMS_TOTAL = 12
+STALE_CACHE_MAX_AGE_SECONDS = 60 * 45
+
+
+@dataclass(frozen=True)
+class RegionProfile:
+    key: str
+    label: str
+    query: str
+    accent: str
+
+
+@dataclass(frozen=True)
+class RegionAnchor:
+    region: str
+    label: str
+    latitude: float
+    longitude: float
+
+
+@dataclass(frozen=True)
+class NewsFeedSource:
+    id: str
+    name: str
+    region: str
+    feed_url: str
+    homepage_url: str
+    source_type: str = "rss"
+    priority: int = 100
+
+
+@dataclass
+class ParsedNewsItem:
+    id: str
+    title: str
+    summary: str
+    url: str
+    source: str
+    feed_name: str
+    feed_region: str
+    homepage_url: str
+    published_at: datetime | None
+
+
+@dataclass
+class CachedRegionFeed:
+    region: str
+    fetched_at: datetime
+    items: list[ParsedNewsItem]
+    sources: list[NewsFeedSource]
+
+
+REGION_PROFILES: dict[str, RegionProfile] = {
+    "americas": RegionProfile(
+        key="americas",
+        label="美洲焦点",
+        query='Americas geopolitics OR Latin America OR "United States" OR Canada',
+        accent="#79d3ff",
+    ),
+    "europe": RegionProfile(
+        key="europe",
+        label="欧洲焦点",
+        query='Europe geopolitics OR EU OR NATO OR "Eastern Europe"',
+        accent="#8fd4ff",
+    ),
+    "middle-east-africa": RegionProfile(
+        key="middle-east-africa",
+        label="中东与非洲焦点",
+        query='"Middle East" OR Africa geopolitics OR Red Sea OR Gulf',
+        accent="#ffb56a",
+    ),
+    "asia-pacific": RegionProfile(
+        key="asia-pacific",
+        label="亚太焦点",
+        query='"Asia Pacific" OR Indo-Pacific OR China OR Japan OR Korea OR ASEAN',
+        accent="#78f2cf",
+    ),
+    "global": RegionProfile(
+        key="global",
+        label="全球焦点",
+        query='"world news" OR geopolitics OR "global affairs"',
+        accent="#d6e6ff",
+    ),
+}
+
+REGION_ANCHORS: dict[str, RegionAnchor] = {
+    "americas": RegionAnchor(
+        region="americas",
+        label="美洲",
+        latitude=37.0902,
+        longitude=-95.7129,
+    ),
+    "europe": RegionAnchor(
+        region="europe",
+        label="欧洲",
+        latitude=50.1109,
+        longitude=8.6821,
+    ),
+    "middle-east-africa": RegionAnchor(
+        region="middle-east-africa",
+        label="中东与非洲",
+        latitude=25.2048,
+        longitude=55.2708,
+    ),
+    "asia-pacific": RegionAnchor(
+        region="asia-pacific",
+        label="亚太",
+        latitude=1.3521,
+        longitude=103.8198,
+    ),
+    "global": RegionAnchor(
+        region="global",
+        label="全球",
+        latitude=20.0,
+        longitude=0.0,
+    ),
+}
+
+
+def _google_news_feed(query: str, *, hl: str, gl: str, ceid: str) -> str:
+    return (
+        "https://news.google.com/rss/search?q="
+        + quote(query, safe="")
+        + f"&hl={hl}&gl={gl}&ceid={ceid}"
+    )
+
+
+NEWS_FEED_SOURCES: tuple[NewsFeedSource, ...] = (
+    NewsFeedSource(
+        id="bbc-world",
+        name="BBC World",
+        region="global",
+        feed_url="https://feeds.bbci.co.uk/news/world/rss.xml",
+        homepage_url="https://www.bbc.com/news/world",
+        priority=10,
+    ),
+    NewsFeedSource(
+        id="dw-top",
+        name="DW Top Stories",
+        region="europe",
+        feed_url="https://rss.dw.com/rdf/rss-en-top",
+        homepage_url="https://www.dw.com/en/top-stories/s-9097",
+        priority=20,
+    ),
+    NewsFeedSource(
+        id="global-scan",
+        name="Global Monitor / World",
+        region="global",
+        feed_url=_google_news_feed(
+            REGION_PROFILES["global"].query,
+            hl="en-US",
+            gl="US",
+            ceid="US:en",
+        ),
+        homepage_url="https://news.google.com/",
+        source_type="aggregated",
+        priority=30,
+    ),
+    NewsFeedSource(
+        id="google-americas",
+        name="Global Monitor / Americas",
+        region="americas",
+        feed_url=_google_news_feed(
+            REGION_PROFILES["americas"].query,
+            hl="en-US",
+            gl="US",
+            ceid="US:en",
+        ),
+        homepage_url="https://news.google.com/",
+        source_type="aggregated",
+        priority=40,
+    ),
+    NewsFeedSource(
+        id="google-europe",
+        name="Global Monitor / Europe",
+        region="europe",
+        feed_url=_google_news_feed(
+            REGION_PROFILES["europe"].query,
+            hl="en-GB",
+            gl="GB",
+            ceid="GB:en",
+        ),
+        homepage_url="https://news.google.com/",
+        source_type="aggregated",
+        priority=40,
+    ),
+    NewsFeedSource(
+        id="google-mea",
+        name="Global Monitor / MEA",
+        region="middle-east-africa",
+        feed_url=_google_news_feed(
+            REGION_PROFILES["middle-east-africa"].query,
+            hl="en-US",
+            gl="US",
+            ceid="US:en",
+        ),
+        homepage_url="https://news.google.com/",
+        source_type="aggregated",
+        priority=40,
+    ),
+    NewsFeedSource(
+        id="google-apac",
+        name="Global Monitor / APAC",
+        region="asia-pacific",
+        feed_url=_google_news_feed(
+            REGION_PROFILES["asia-pacific"].query,
+            hl="en-SG",
+            gl="SG",
+            ceid="SG:en",
+        ),
+        homepage_url="https://news.google.com/",
+        source_type="aggregated",
+        priority=40,
+    ),
+)
+
+
+_REGION_CACHE: dict[str, CachedRegionFeed] = {}
+
+
+def determine_focus_region(lat: float | None, lon: float | None) -> str:
+    if lat is None or lon is None:
+        return "global"
+    if -170 <= lon <= -30:
+        return "americas"
+    if -30 < lon <= 45:
+        return "europe" if lat >= 30 else "middle-east-africa"
+    if 45 < lon <= 150:
+        return "middle-east-africa" if lat < 10 else "asia-pacific"
+    return "asia-pacific"
+
+
+def get_region_profile(region: str) -> RegionProfile:
+    return REGION_PROFILES.get(region, REGION_PROFILES["global"])
+
+
+def get_region_anchor(region: str) -> RegionAnchor:
+    return REGION_ANCHORS.get(region, REGION_ANCHORS["global"])
+
+
+def get_sources_for_region(region: str) -> list[NewsFeedSource]:
+    return sorted(
+        [source for source in NEWS_FEED_SOURCES if source.region in {"global", region}],
+        key=lambda source: (source.priority, source.name),
+    )
+
+
+def _strip_html(value: str) -> str:
+    if not value:
+        return ""
+    soup = BeautifulSoup(value, "html.parser")
+    return re.sub(r"\s+", " ", soup.get_text(" ", strip=True)).strip()
+
+
+def _truncate(value: str, limit: int = 180) -> str:
+    text = value.strip()
+    if len(text) <= limit:
+        return text
+    return text[: limit - 1].rstrip() + "…"
+
+
+def _normalize_source_name(raw: str, fallback: str) -> str:
+    text = html.unescape((raw or "").strip())
+    if " - " in text:
+        return text.split(" - ")[-1].strip() or fallback
+    return text or fallback
+
+
+def _parse_datetime(raw: str | None) -> datetime | None:
+    if not raw:
+        return None
+    text = raw.strip()
+    if not text:
+        return None
+
+    for parser in (
+        lambda value: parsedate_to_datetime(value),
+        lambda value: datetime.fromisoformat(value.replace("Z", "+00:00")),
+    ):
+        try:
+            parsed = parser(text)
+            if parsed.tzinfo is None:
+                parsed = parsed.replace(tzinfo=UTC)
+            return parsed.astimezone(UTC)
+        except Exception:
+            continue
+    return None
+
+
+def _extract_item_text(element: ET.Element, *names: str) -> str:
+    for name in names:
+        node = element.find(name)
+        if node is not None and node.text:
+            return node.text.strip()
+    return ""
+
+
+def _parse_feed_entries(xml_text: str, source: NewsFeedSource) -> list[ParsedNewsItem]:
+    root = ET.fromstring(xml_text)
+    items: list[ParsedNewsItem] = []
+
+    rss_items = root.findall("./channel/item")
+    atom_entries = root.findall("{http://www.w3.org/2005/Atom}entry")
+    nodes = rss_items or atom_entries
+
+    for node in nodes[:MAX_ITEMS_PER_SOURCE]:
+        if node.tag.endswith("entry"):
+            title = _extract_item_text(node, "{http://www.w3.org/2005/Atom}title")
+            summary = _extract_item_text(
+                node,
+                "{http://www.w3.org/2005/Atom}summary",
+                "{http://www.w3.org/2005/Atom}content",
+            )
+            link_node = node.find("{http://www.w3.org/2005/Atom}link")
+            link = link_node.get("href", "").strip() if link_node is not None else ""
+            published = _extract_item_text(
+                node,
+                "{http://www.w3.org/2005/Atom}updated",
+                "{http://www.w3.org/2005/Atom}published",
+            )
+        else:
+            title = _extract_item_text(node, "title")
+            summary = _extract_item_text(node, "description", "content")
+            link = _extract_item_text(node, "link")
+            published = _extract_item_text(node, "pubDate", "published", "updated")
+
+        clean_title = html.unescape(title).strip()
+        clean_summary = _truncate(_strip_html(summary), 180)
+        if not clean_title or not link:
+            continue
+
+        item_source = _normalize_source_name(clean_title, source.name)
+        display_title = clean_title
+        if source.source_type == "aggregated" and " - " in clean_title:
+            parts = clean_title.rsplit(" - ", 1)
+            display_title = parts[0].strip()
+            item_source = _normalize_source_name(parts[1], source.name)
+
+        items.append(
+            ParsedNewsItem(
+                id=f"{source.id}:{hashlib.sha1(link.encode('utf-8')).hexdigest()[:12]}",
+                title=display_title,
+                summary=clean_summary,
+                url=link,
+                source=item_source,
+                feed_name=source.name,
+                feed_region=source.region,
+                homepage_url=source.homepage_url,
+                published_at=_parse_datetime(published),
+            )
+        )
+
+    return items
+
+
+def _serialize_sources(sources: list[NewsFeedSource]) -> list[dict[str, Any]]:
+    return [
+        {
+            "id": source.id,
+            "name": source.name,
+            "region": source.region,
+            "homepage_url": source.homepage_url,
+        }
+        for source in sources
+    ]
+
+
+def _serialize_item(item: ParsedNewsItem, *, active_region: str) -> dict[str, Any]:
+    published_at = item.published_at
+    anchor = get_region_anchor(item.feed_region)
+    return {
+        "id": item.id,
+        "title": item.title,
+        "summary": item.summary,
+        "url": item.url,
+        "source": item.source,
+        "feed_name": item.feed_name,
+        "region": item.feed_region,
+        "homepage_url": item.homepage_url,
+        "published_at": published_at.isoformat().replace("+00:00", "Z") if published_at else None,
+        "latitude": anchor.latitude,
+        "longitude": anchor.longitude,
+        "location_label": anchor.label,
+        "location_inferred": True,
+        "is_focus_match": item.feed_region == active_region,
+    }
+
+
+def _build_payload(
+    *,
+    lat: float | None,
+    lon: float | None,
+    active_region: str,
+    items: list[ParsedNewsItem],
+    sources: list[NewsFeedSource],
+    errors: list[str],
+    stale: bool,
+    generated_at: datetime | None = None,
+) -> dict[str, Any]:
+    profile = get_region_profile(active_region)
+    timestamp = generated_at or datetime.now(UTC)
+    return {
+        "generated_at": timestamp.isoformat().replace("+00:00", "Z"),
+        "focus": {
+            "lat": lat,
+            "lon": lon,
+            "region": active_region,
+            "label": profile.label,
+            "accent": profile.accent,
+        },
+        "sources": _serialize_sources(sources),
+        "items": [_serialize_item(item, active_region=active_region) for item in items],
+        "errors": errors,
+        "stale": stale,
+    }
+
+
+def _rank_and_trim_items(items: list[ParsedNewsItem], *, active_region: str) -> list[ParsedNewsItem]:
+    deduped: dict[str, ParsedNewsItem] = {}
+    for item in items:
+        key = item.url.strip() or item.title.strip().lower()
+        if key not in deduped:
+            deduped[key] = item
+
+    return sorted(
+        deduped.values(),
+        key=lambda item: (
+            item.feed_region != active_region,
+            item.published_at is None,
+            -(item.published_at.timestamp() if item.published_at else 0),
+            item.feed_name,
+        ),
+    )[:MAX_ITEMS_TOTAL]
+
+
+def _get_cached_region_feed(region: str) -> CachedRegionFeed | None:
+    cached = _REGION_CACHE.get(region)
+    if not cached:
+        return None
+    age_seconds = (datetime.now(UTC) - cached.fetched_at).total_seconds()
+    if age_seconds > STALE_CACHE_MAX_AGE_SECONDS:
+        return None
+    return cached
+
+
+def _store_region_cache(region: str, *, items: list[ParsedNewsItem], sources: list[NewsFeedSource]) -> None:
+    _REGION_CACHE[region] = CachedRegionFeed(
+        region=region,
+        fetched_at=datetime.now(UTC),
+        items=list(items),
+        sources=list(sources),
+    )
+
+
+async def _fetch_source(
+    client: httpx.AsyncClient,
+    source: NewsFeedSource,
+) -> tuple[NewsFeedSource, list[ParsedNewsItem], str | None]:
+    try:
+        response = await client.get(source.feed_url)
+        response.raise_for_status()
+        return source, _parse_feed_entries(response.text, source), None
+    except Exception as exc:
+        return source, [], str(exc)
+
+
+async def get_earth_news_payload(lat: float | None = None, lon: float | None = None) -> dict[str, Any]:
+    active_region = determine_focus_region(lat, lon)
+    sources = get_sources_for_region(active_region)
+    errors: list[str] = []
+
+    async with httpx.AsyncClient(
+        timeout=REQUEST_TIMEOUT,
+        follow_redirects=True,
+        headers={"User-Agent": USER_AGENT},
+    ) as client:
+        results = await asyncio.gather(*(_fetch_source(client, source) for source in sources))
+
+    fetched_items: list[ParsedNewsItem] = []
+    for source, items, error in results:
+        if error:
+            errors.append(f"{source.name}: {error}")
+            continue
+        fetched_items.extend(items)
+
+    ranked_items = _rank_and_trim_items(fetched_items, active_region=active_region)
+    if ranked_items:
+        _store_region_cache(active_region, items=ranked_items, sources=sources)
+        return _build_payload(
+            lat=lat,
+            lon=lon,
+            active_region=active_region,
+            items=ranked_items,
+            sources=sources,
+            errors=errors,
+            stale=False,
+        )
+
+    cached = _get_cached_region_feed(active_region)
+    if cached:
+        return _build_payload(
+            lat=lat,
+            lon=lon,
+            active_region=active_region,
+            items=cached.items,
+            sources=cached.sources,
+            errors=errors,
+            stale=True,
+            generated_at=cached.fetched_at,
+        )
+
+    return _build_payload(
+        lat=lat,
+        lon=lon,
+        active_region=active_region,
+        items=[],
+        sources=sources,
+        errors=errors,
+        stale=False,
+    )

backend/app/services/persistent_logs.py

@@ -0,0 +1,86 @@
+from __future__ import annotations
+
+from typing import Any
+
+from app.core.logging import get_logger, sanitize_log_value
+from app.core.request_context import get_request_id
+from app.db.session import async_session_factory
+from app.models.system_log import AuditLog, SystemLog
+
+logger = get_logger(__name__)
+
+
+async def record_system_log(
+    *,
+    source: str,
+    level: str,
+    message: str,
+    service: str | None = None,
+    module: str | None = None,
+    event: str | None = None,
+    request_id: str | None = None,
+    trace_id: str | None = None,
+    user_id: int | None = None,
+    category: str | None = None,
+    context: dict[str, Any] | None = None,
+) -> None:
+    try:
+        async with async_session_factory() as session:
+            session.add(
+                SystemLog(
+                    source=source,
+                    service=service,
+                    module=module,
+                    event=event,
+                    level=level.lower(),
+                    message=str(sanitize_log_value(message)),
+                    request_id=request_id or get_request_id(),
+                    trace_id=trace_id,
+                    user_id=user_id,
+                    category=category,
+                    context=sanitize_log_value(context or {}),
+                )
+            )
+            await session.commit()
+    except Exception:
+        logger.exception_event(
+            "Failed to persist system log",
+            event="system_log.persist.failed",
+            context={"event_name": event, "source": source},
+        )
+
+
+async def record_audit_log(
+    *,
+    action: str,
+    actor_id: int | None = None,
+    actor_name: str | None = None,
+    target_type: str | None = None,
+    target_id: str | None = None,
+    result: str | None = None,
+    request_id: str | None = None,
+    ip: str | None = None,
+    details: dict[str, Any] | None = None,
+) -> None:
+    try:
+        async with async_session_factory() as session:
+            session.add(
+                AuditLog(
+                    actor_id=actor_id,
+                    actor_name=actor_name,
+                    action=action,
+                    target_type=target_type,
+                    target_id=target_id,
+                    result=result,
+                    request_id=request_id or get_request_id(),
+                    ip=ip,
+                    details=sanitize_log_value(details or {}),
+                )
+            )
+            await session.commit()
+    except Exception:
+        logger.exception_event(
+            "Failed to persist audit log",
+            event="audit_log.persist.failed",
+            context={"action": action},
+        )

backend/app/services/playground_chat_service.py

@@ -0,0 +1,694 @@
+from __future__ import annotations
+
+import asyncio
+from collections.abc import Sequence
+from datetime import UTC, datetime
+from time import perf_counter
+from uuid import uuid4
+
+from fastapi import HTTPException, status
+from sqlalchemy import func, select, update
+from sqlalchemy.ext.asyncio import AsyncSession
+
+from app.db.session import async_session_factory
+from app.models.playground_message import PlaygroundMessage
+from app.models.playground_session import PlaygroundSession
+from app.schemas.ai import (
+    PlaygroundMessageEditRequest,
+    PlaygroundMessageActionResponse,
+    PlaygroundMessageCreateRequest,
+    PlaygroundMessageRecord,
+    PlaygroundMessageResendRequest,
+    PlaygroundMessageStopRequest,
+    PlaygroundSessionResponse,
+    PlaygroundSessionState,
+    PlaygroundSessionUpsertRequest,
+    PlaygroundThreadResponse,
+    SituationalAnalysisRequest,
+)
+from app.services.ai_client import AIProviderClient
+from app.services.playground_session_store import _to_response as session_to_response
+from app.services.playground_session_store import upsert_playground_session
+
+STREAM_CHUNK_SIZE = 24
+STREAM_INTERVAL_SECONDS = 0.08
+THINKING_PREVIEW_SECONDS = 2.6
+
+
+class _ActiveRun:
+    def __init__(self, task: asyncio.Task[None]) -> None:
+        self.task = task
+        self.stop_requested = asyncio.Event()
+
+
+_ACTIVE_RUNS: dict[str, _ActiveRun] = {}
+
+
+async def _get_session_by_key(
+    db: AsyncSession,
+    *,
+    user_id: int,
+    session_key: str,
+) -> PlaygroundSession | None:
+    result = await db.execute(
+        select(PlaygroundSession).where(
+            PlaygroundSession.user_id == user_id,
+            PlaygroundSession.session_key == session_key,
+        )
+    )
+    return result.scalar_one_or_none()
+
+
+async def _require_session(
+    db: AsyncSession,
+    *,
+    user_id: int,
+    session_key: str,
+) -> PlaygroundSession:
+    session = await _get_session_by_key(db, user_id=user_id, session_key=session_key)
+    if session is None:
+        raise HTTPException(status_code=status.HTTP_404_NOT_FOUND, detail="Playground session not found")
+    return session
+
+
+async def _require_visible_message(
+    db: AsyncSession,
+    *,
+    user_id: int,
+    public_id: str,
+    role: str | None = None,
+) -> PlaygroundMessage:
+    conditions = [
+        PlaygroundMessage.user_id == user_id,
+        PlaygroundMessage.public_id == public_id,
+        PlaygroundMessage.is_visible.is_(True),
+    ]
+    if role is not None:
+        conditions.append(PlaygroundMessage.role == role)
+
+    result = await db.execute(select(PlaygroundMessage).where(*conditions))
+    message = result.scalar_one_or_none()
+    if message is None:
+        detail = "User message not found" if role == "user" else "Playground message not found"
+        raise HTTPException(status_code=status.HTTP_404_NOT_FOUND, detail=detail)
+    return message
+
+
+def _message_to_record(message: PlaygroundMessage, parent_public_id: str | None = None) -> PlaygroundMessageRecord:
+    return PlaygroundMessageRecord(
+        id=message.public_id,
+        role=message.role,
+        kind=message.kind,
+        status=message.status,
+        title=message.title,
+        content=message.content or "",
+        thinking_content=message.thinking_content or "",
+        meta=list(message.meta or []),
+        markdown=message.role != "system",
+        provider=message.provider,
+        model=message.model,
+        request_id=message.request_id,
+        raw_response=dict(message.raw_response or {}),
+        content_blocks=list(message.content_blocks or []),
+        text_blocks=list(message.text_blocks or []),
+        thinking_blocks=list(message.thinking_blocks or []),
+        parent_message_id=parent_public_id,
+        created_at=message.created_at.isoformat(),
+        updated_at=message.updated_at.isoformat(),
+    )
+
+
+async def _ensure_session(
+    db: AsyncSession,
+    *,
+    user_id: int,
+    session_key: str,
+    title: str,
+    state: PlaygroundSessionState | None = None,
+) -> PlaygroundSession:
+    result = await db.execute(
+        select(PlaygroundSession).where(
+            PlaygroundSession.user_id == user_id,
+            PlaygroundSession.session_key == session_key,
+        )
+    )
+    session = result.scalar_one_or_none()
+    if session is not None:
+        if title:
+            session.title = title[:200]
+        if state is not None:
+            session.state = state.model_dump(mode="json")
+        await db.flush()
+        await db.refresh(session)
+        return session
+
+    payload = PlaygroundSessionUpsertRequest(
+        session_key=session_key,
+        title=title[:200],
+        state=state or PlaygroundSessionState(title=title[:200]),
+    )
+    await upsert_playground_session(db, user_id=user_id, payload=payload)
+    result = await db.execute(
+        select(PlaygroundSession).where(
+            PlaygroundSession.user_id == user_id,
+            PlaygroundSession.session_key == session_key,
+        )
+    )
+    return result.scalar_one()
+
+
+async def _list_visible_messages(
+    db: AsyncSession,
+    *,
+    session_id: int,
+) -> list[PlaygroundMessage]:
+    result = await db.execute(
+        select(PlaygroundMessage)
+        .where(
+            PlaygroundMessage.session_id == session_id,
+            PlaygroundMessage.is_visible.is_(True),
+        )
+        .order_by(PlaygroundMessage.sort_order.asc(), PlaygroundMessage.id.asc())
+    )
+    return list(result.scalars().all())
+
+
+async def _build_thread_response(
+    db: AsyncSession,
+    *,
+    session: PlaygroundSession,
+) -> PlaygroundThreadResponse:
+    messages = await _list_visible_messages(db, session_id=session.id)
+    id_map = {item.id: item.public_id for item in messages}
+    return PlaygroundThreadResponse(
+        session=session_to_response(session),
+        messages=[_message_to_record(item, id_map.get(item.parent_message_id)) for item in messages],
+    )
+
+
+async def get_thread(
+    db: AsyncSession,
+    *,
+    user_id: int,
+    session_key: str,
+) -> PlaygroundThreadResponse | None:
+    session = await _get_session_by_key(db, user_id=user_id, session_key=session_key)
+    if session is None:
+        return None
+    return await _build_thread_response(db, session=session)
+
+
+async def _build_action_response(
+    db: AsyncSession,
+    *,
+    session: PlaygroundSession,
+    active_message_id: str | None = None,
+) -> PlaygroundMessageActionResponse:
+    thread = await _build_thread_response(db, session=session)
+    return PlaygroundMessageActionResponse(
+        session=thread.session,
+        messages=thread.messages,
+        active_message_id=active_message_id,
+    )
+
+
+def _collect_constraints(raw_constraints: str) -> list[str]:
+    return [item.strip() for item in raw_constraints.split("\n") if item.strip()]
+
+
+async def _next_sort_order(db: AsyncSession, session_id: int) -> int:
+    result = await db.execute(
+        select(func.max(PlaygroundMessage.sort_order)).where(PlaygroundMessage.session_id == session_id)
+    )
+    current = result.scalar_one_or_none()
+    return int(current or 0)
+
+
+async def _set_session_state(
+    db: AsyncSession,
+    *,
+    session: PlaygroundSession,
+    payload: PlaygroundMessageCreateRequest,
+) -> PlaygroundSession:
+    session.state = PlaygroundSessionState(
+        messages=[],
+        selectedPresetKey=payload.selected_preset_key,
+        title=payload.title,
+        objective=payload.objective,
+        constraints=payload.constraints,
+        inputValue="",
+        analysis=None,
+        latestAnalysisMessageId=None,
+        analysisMeta={},
+        helpExpanded=payload.help_expanded,
+    ).model_dump(mode="json")
+    session.title = payload.title[:200]
+    await db.flush()
+    await db.refresh(session)
+    return session
+
+
+def _spawn_assistant_run(
+    *,
+    user_id: int,
+    session_id: int,
+    session_key: str,
+    user_message_id: int,
+    assistant_message_id: int,
+    assistant_public_id: str,
+    payload: PlaygroundMessageCreateRequest,
+    provider_client: AIProviderClient,
+) -> None:
+    task = asyncio.create_task(
+        _run_assistant_message(
+            user_id=user_id,
+            session_id=session_id,
+            session_key=session_key,
+            user_message_id=user_message_id,
+            assistant_message_id=assistant_message_id,
+            payload=payload,
+            provider_client=provider_client,
+        )
+    )
+    _ACTIVE_RUNS[assistant_public_id] = _ActiveRun(task)
+
+
+async def create_turn(
+    db: AsyncSession,
+    *,
+    user_id: int,
+    payload: PlaygroundMessageCreateRequest,
+    provider_client: AIProviderClient,
+) -> PlaygroundMessageActionResponse:
+    session = await _ensure_session(
+        db,
+        user_id=user_id,
+        session_key=payload.session_key,
+        title=payload.title,
+        state=PlaygroundSessionState(
+            selectedPresetKey=payload.selected_preset_key,
+            title=payload.title,
+            objective=payload.objective,
+            constraints=payload.constraints,
+            inputValue="",
+            helpExpanded=payload.help_expanded,
+        ),
+    )
+    session = await _set_session_state(db, session=session, payload=payload)
+    base_order = await _next_sort_order(db, session.id)
+
+    user_message = PlaygroundMessage(
+        public_id=uuid4().hex,
+        session_id=session.id,
+        user_id=user_id,
+        role="user",
+        kind="message",
+        status="done",
+        title=payload.selected_preset_key,
+        content=payload.input,
+        meta=[payload.title],
+        sort_order=base_order + 10,
+    )
+    assistant_message = PlaygroundMessage(
+        public_id=uuid4().hex,
+        session_id=session.id,
+        user_id=user_id,
+        parent_message_id=None,
+        role="assistant",
+        kind="thinking",
+        status="pending",
+        title="AI 回应",
+        content="",
+        thinking_content="",
+        meta=[],
+        sort_order=base_order + 20,
+    )
+    db.add(user_message)
+    await db.flush()
+    assistant_message.parent_message_id = user_message.id
+    db.add(assistant_message)
+    await db.flush()
+    await db.refresh(user_message)
+    await db.refresh(assistant_message)
+    await db.commit()
+    await db.refresh(session)
+    await db.refresh(user_message)
+    await db.refresh(assistant_message)
+
+    _spawn_assistant_run(
+        user_id=user_id,
+        session_id=session.id,
+        session_key=payload.session_key,
+        user_message_id=user_message.id,
+        assistant_message_id=assistant_message.id,
+        assistant_public_id=assistant_message.public_id,
+        payload=payload,
+        provider_client=provider_client,
+    )
+
+    return await _build_action_response(
+        db,
+        session=session,
+        active_message_id=assistant_message.public_id,
+    )
+
+
+async def _create_assistant_retry_turn(
+    db: AsyncSession,
+    *,
+    user_id: int,
+    session: PlaygroundSession,
+    user_message: PlaygroundMessage,
+    payload: PlaygroundMessageCreateRequest,
+    provider_client: AIProviderClient,
+) -> PlaygroundMessageActionResponse:
+    base_order = await _next_sort_order(db, session.id)
+    assistant_message = PlaygroundMessage(
+        public_id=uuid4().hex,
+        session_id=session.id,
+        user_id=user_id,
+        parent_message_id=user_message.id,
+        role="assistant",
+        kind="thinking",
+        status="pending",
+        title="AI 回应",
+        content="",
+        thinking_content="",
+        meta=[],
+        sort_order=base_order + 10,
+    )
+    db.add(assistant_message)
+    await db.flush()
+    await db.commit()
+    await db.refresh(session)
+    await db.refresh(assistant_message)
+
+    _spawn_assistant_run(
+        user_id=user_id,
+        session_id=session.id,
+        session_key=payload.session_key,
+        user_message_id=user_message.id,
+        assistant_message_id=assistant_message.id,
+        assistant_public_id=assistant_message.public_id,
+        payload=payload,
+        provider_client=provider_client,
+    )
+
+    return await _build_action_response(
+        db,
+        session=session,
+        active_message_id=assistant_message.public_id,
+    )
+
+
+async def stop_message(
+    db: AsyncSession,
+    *,
+    user_id: int,
+    payload: PlaygroundMessageStopRequest,
+) -> PlaygroundMessageActionResponse:
+    session = await _require_session(db, user_id=user_id, session_key=payload.session_key)
+    message = await _require_visible_message(db, user_id=user_id, public_id=payload.message_id)
+
+    if message.status not in {"pending", "thinking", "answering"}:
+        return await _build_action_response(db, session=session)
+
+    active_run = _ACTIVE_RUNS.get(message.public_id)
+    if active_run is not None:
+        active_run.stop_requested.set()
+        active_run.task.cancel()
+
+    message.status = "stopped"
+    if "已手动停止生成" not in (message.meta or []):
+        message.meta = [*(message.meta or []), "已手动停止生成"]
+    await db.flush()
+    await db.commit()
+    await db.refresh(message)
+
+    return await _build_action_response(db, session=session)
+
+
+async def resend_turn(
+    db: AsyncSession,
+    *,
+    user_id: int,
+    payload: PlaygroundMessageResendRequest,
+    provider_client: AIProviderClient,
+) -> PlaygroundMessageActionResponse:
+    session = await _require_session(db, user_id=user_id, session_key=payload.session_key)
+    user_message = await _require_visible_message(
+        db,
+        user_id=user_id,
+        public_id=payload.user_message_id,
+        role="user",
+    )
+
+    later_messages = await db.execute(
+        select(PlaygroundMessage).where(
+            PlaygroundMessage.session_id == session.id,
+            PlaygroundMessage.sort_order > user_message.sort_order,
+            PlaygroundMessage.is_visible.is_(True),
+        )
+    )
+    for item in later_messages.scalars().all():
+        item.is_visible = False
+        if item.status in {"pending", "thinking", "answering"}:
+            active_run = _ACTIVE_RUNS.get(item.public_id)
+            if active_run is not None:
+                active_run.stop_requested.set()
+                active_run.task.cancel()
+    await db.flush()
+    await db.commit()
+
+    session_state = PlaygroundSessionState.model_validate(session.state or {})
+    create_payload = PlaygroundMessageCreateRequest(
+        session_key=payload.session_key,
+        title=session_state.title or session.title,
+        objective=session_state.objective or "继续当前对话",
+        constraints=session_state.constraints or "",
+        input=user_message.content,
+        selected_preset_key=session_state.selectedPresetKey or "bgp-brief",
+        help_expanded=session_state.helpExpanded,
+    )
+    return await _create_assistant_retry_turn(
+        db,
+        user_id=user_id,
+        session=session,
+        user_message=user_message,
+        payload=create_payload,
+        provider_client=provider_client,
+    )
+
+
+async def edit_user_message(
+    db: AsyncSession,
+    *,
+    user_id: int,
+    payload: PlaygroundMessageEditRequest,
+) -> PlaygroundMessageActionResponse:
+    session = await _require_session(db, user_id=user_id, session_key=payload.session_key)
+    user_message = await _require_visible_message(
+        db,
+        user_id=user_id,
+        public_id=payload.user_message_id,
+        role="user",
+    )
+
+    user_message.content = payload.content.strip()
+    await db.flush()
+    await db.commit()
+    await db.refresh(user_message)
+
+    return await _build_action_response(db, session=session)
+
+
+async def _append_meta_if_missing(db: AsyncSession, message_id: int, meta_line: str) -> None:
+    result = await db.execute(select(PlaygroundMessage).where(PlaygroundMessage.id == message_id))
+    message = result.scalar_one_or_none()
+    if message is None:
+        return
+    if meta_line not in (message.meta or []):
+        message.meta = [*(message.meta or []), meta_line]
+        await db.flush()
+
+
+async def _should_stop(message_public_id: str) -> bool:
+    active_run = _ACTIVE_RUNS.get(message_public_id)
+    return active_run.stop_requested.is_set() if active_run is not None else False
+
+
+async def _mark_message_state(
+    db: AsyncSession,
+    *,
+    message_id: int,
+    **updates,
+) -> PlaygroundMessage:
+    result = await db.execute(select(PlaygroundMessage).where(PlaygroundMessage.id == message_id))
+    message = result.scalar_one()
+    for key, value in updates.items():
+        setattr(message, key, value)
+    await db.flush()
+    await db.refresh(message)
+    return message
+
+
+def _build_conversation_history(messages: Sequence[PlaygroundMessage], current_user_message_id: int) -> list[dict]:
+    history: list[dict] = []
+    for item in messages:
+        if item.id >= current_user_message_id:
+            break
+        if item.role == "system":
+            continue
+        history.append(
+            {
+                "role": item.role,
+                "kind": item.kind or "message",
+                "title": item.title,
+                "content": item.content or "",
+            }
+        )
+    return history[-8:]
+
+
+async def _run_assistant_message(
+    *,
+    user_id: int,
+    session_id: int,
+    session_key: str,
+    user_message_id: int,
+    assistant_message_id: int,
+    payload: PlaygroundMessageCreateRequest,
+    provider_client: AIProviderClient,
+) -> None:
+    request_id = str(uuid4())
+    started_at = perf_counter()
+    assistant_public_id: str | None = None
+    try:
+        async with async_session_factory() as db:
+            session = await db.get(PlaygroundSession, session_id)
+            user_message = await db.get(PlaygroundMessage, user_message_id)
+            assistant_message = await db.get(PlaygroundMessage, assistant_message_id)
+            if session is None or user_message is None or assistant_message is None:
+                return
+            assistant_public_id = assistant_message.public_id
+
+            visible_messages = await _list_visible_messages(db, session_id=session_id)
+            conversation_history = _build_conversation_history(visible_messages, user_message_id)
+
+            request_payload = SituationalAnalysisRequest(
+                title=payload.title,
+                objective=payload.objective,
+                observations=[item.strip() for item in payload.input.split("\n") if item.strip()],
+                constraints=_collect_constraints(payload.constraints),
+                context={
+                    "source": "playground",
+                    "preset": payload.selected_preset_key,
+                    "conversation_history": conversation_history,
+                    "history_size": len(conversation_history),
+                },
+                thinking={"type": "enabled"},
+            )
+
+        analysis = await provider_client.analyze(request_payload, request_id=request_id)
+
+        async with async_session_factory() as db:
+            assistant_message = await _mark_message_state(
+                db,
+                message_id=assistant_message_id,
+                status="thinking" if analysis.thinking_blocks else "answering",
+                title=f"{analysis.provider} / {analysis.model}",
+                provider=analysis.provider,
+                model=analysis.model,
+                request_id=request_id,
+                raw_response=analysis.raw_response,
+                content_blocks=[item.model_dump(mode="json") for item in analysis.content_blocks],
+                text_blocks=analysis.text_blocks,
+                thinking_blocks=analysis.thinking_blocks,
+                thinking_content="\n\n".join(analysis.thinking_blocks).strip(),
+            )
+            session = await db.get(PlaygroundSession, session_id)
+            if session is not None:
+                session_state = PlaygroundSessionState.model_validate(session.state or {})
+                session.state = session_state.model_copy(
+                    update={
+                        "latestAnalysisMessageId": assistant_message.public_id,
+                        "analysis": analysis.model_dump(mode="json"),
+                    }
+                ).model_dump(mode="json")
+                await db.flush()
+            await db.commit()
+
+        if assistant_public_id and analysis.thinking_blocks:
+            await asyncio.sleep(THINKING_PREVIEW_SECONDS)
+            if await _should_stop(assistant_public_id):
+                return
+
+        content = analysis.content or ""
+        cursor = 0
+        while cursor < len(content):
+            if assistant_public_id and await _should_stop(assistant_public_id):
+                return
+            cursor = min(len(content), cursor + STREAM_CHUNK_SIZE)
+            async with async_session_factory() as db:
+                await _mark_message_state(
+                    db,
+                    message_id=assistant_message_id,
+                    status="answering",
+                    content=content[:cursor],
+                )
+                await db.commit()
+            await asyncio.sleep(STREAM_INTERVAL_SECONDS)
+
+        duration_ms = round((perf_counter() - started_at) * 1000)
+        async with async_session_factory() as db:
+            assistant_message = await _mark_message_state(
+                db,
+                message_id=assistant_message_id,
+                status="done",
+                content=content,
+                meta=[
+                    f"Request ID: {request_id}",
+                    f"耗时: {duration_ms} ms",
+                    f"完成时间: {datetime.now(UTC).astimezone().isoformat()}",
+                ],
+            )
+            session = await db.get(PlaygroundSession, session_id)
+            if session is not None:
+                session_state = PlaygroundSessionState.model_validate(session.state or {})
+                session.state = session_state.model_copy(
+                    update={
+                        "latestAnalysisMessageId": assistant_message.public_id,
+                        "analysis": analysis.model_dump(mode="json"),
+                        "analysisMeta": {
+                            "requestId": request_id,
+                            "durationMs": duration_ms,
+                            "completedAt": datetime.now().isoformat(),
+                        },
+                    }
+                ).model_dump(mode="json")
+                await db.flush()
+            await db.commit()
+    except asyncio.CancelledError:
+        async with async_session_factory() as db:
+            result = await db.execute(select(PlaygroundMessage).where(PlaygroundMessage.id == assistant_message_id))
+            message = result.scalar_one_or_none()
+            if message is not None and message.status in {"pending", "thinking", "answering"}:
+                message.status = "stopped"
+                if "已手动停止生成" not in (message.meta or []):
+                    message.meta = [*(message.meta or []), "已手动停止生成"]
+                await db.flush()
+                await db.commit()
+        raise
+    except Exception as exc:
+        async with async_session_factory() as db:
+            result = await db.execute(select(PlaygroundMessage).where(PlaygroundMessage.id == assistant_message_id))
+            message = result.scalar_one_or_none()
+            if message is not None:
+                message.status = "error"
+                message.content = message.content or "分析失败，请检查 AI Provider 配置或稍后再试。"
+                message.meta = [*(message.meta or []), f"错误: {type(exc).__name__}"]
+                await db.flush()
+                await db.commit()
+    finally:
+        if assistant_public_id:
+            _ACTIVE_RUNS.pop(assistant_public_id, None)

backend/app/services/playground_session_store.py

@@ -0,0 +1,72 @@
+from __future__ import annotations
+
+from sqlalchemy import select
+from sqlalchemy.ext.asyncio import AsyncSession
+
+from app.models.playground_session import PlaygroundSession
+from app.schemas.ai import (
+    PlaygroundSessionResponse,
+    PlaygroundSessionState,
+    PlaygroundSessionUpsertRequest,
+)
+
+
+def _to_response(record: PlaygroundSession) -> PlaygroundSessionResponse:
+    return PlaygroundSessionResponse(
+        id=str(record.id),
+        session_key=record.session_key,
+        title=record.title,
+        state=PlaygroundSessionState.model_validate(record.state or {}),
+        created_at=record.created_at.isoformat(),
+        updated_at=record.updated_at.isoformat(),
+    )
+
+
+async def get_playground_session(
+    db: AsyncSession,
+    *,
+    user_id: int,
+    session_key: str = "default",
+) -> PlaygroundSessionResponse | None:
+    result = await db.execute(
+        select(PlaygroundSession).where(
+            PlaygroundSession.user_id == user_id,
+            PlaygroundSession.session_key == session_key,
+        )
+    )
+    record = result.scalar_one_or_none()
+    if record is None:
+        return None
+    return _to_response(record)
+
+
+async def upsert_playground_session(
+    db: AsyncSession,
+    *,
+    user_id: int,
+    payload: PlaygroundSessionUpsertRequest,
+) -> PlaygroundSessionResponse:
+    result = await db.execute(
+        select(PlaygroundSession).where(
+            PlaygroundSession.user_id == user_id,
+            PlaygroundSession.session_key == payload.session_key,
+        )
+    )
+    record = result.scalar_one_or_none()
+    title = (payload.title or payload.state.title or "Playground 会话").strip()[:200] or "Playground 会话"
+
+    if record is None:
+        record = PlaygroundSession(
+            user_id=user_id,
+            session_key=payload.session_key,
+            title=title,
+            state=payload.state.model_dump(mode="json"),
+        )
+        db.add(record)
+    else:
+        record.title = title
+        record.state = payload.state.model_dump(mode="json")
+
+    await db.flush()
+    await db.refresh(record)
+    return _to_response(record)

backend/app/services/scheduler.py

@@ -1,7 +1,6 @@
 """Task Scheduler for running collection jobs."""
 
 import asyncio
-import logging
 from datetime import UTC, datetime, timedelta
 from typing import Any, Dict, Optional
 
@@ -9,16 +8,41 @@ from apscheduler.schedulers.asyncio import AsyncIOScheduler
 from apscheduler.triggers.interval import IntervalTrigger
 from sqlalchemy import select
 
+from app.core.logging import get_logger
 from app.db.session import async_session_factory
 from app.core.time import to_iso8601_utc
 from app.models.datasource import DataSource
 from app.models.task import CollectionTask
 from app.services.collectors.registry import collector_registry
 
-logger = logging.getLogger(__name__)
+logger = get_logger(__name__)
 
 scheduler = AsyncIOScheduler()
 RUNNING_TASK_GUARD_TIMEOUT_MINUTES = 90
+RUNNING_COLLECTOR_TASKS: dict[str, asyncio.Task[Any]] = {}
+
+
+def _collector_task_name(collector_name: str) -> str:
+    return f"collector:{collector_name}"
+
+
+def get_running_collector_task(collector_name: str) -> asyncio.Task[Any] | None:
+    task = RUNNING_COLLECTOR_TASKS.get(collector_name)
+    if task is not None and not task.done():
+        return task
+
+    if task is not None and task.done():
+        RUNNING_COLLECTOR_TASKS.pop(collector_name, None)
+
+    target_name = _collector_task_name(collector_name)
+    for candidate in asyncio.all_tasks():
+        if candidate.done():
+            continue
+        if candidate.get_name() == target_name:
+            RUNNING_COLLECTOR_TASKS[collector_name] = candidate
+            return candidate
+
+    return None
 
 
 async def _update_next_run_at(datasource: DataSource, session) -> None:
@@ -30,7 +54,11 @@ async def _update_next_run_at(datasource: DataSource, session) -> None:
 async def _apply_datasource_schedule(datasource: DataSource, session) -> None:
     collector = collector_registry.get(datasource.source)
     if not collector:
-        logger.warning("Collector not found for datasource %s", datasource.source)
+        logger.warning_event(
+            "Collector not found for datasource",
+            event="collector.schedule.collector_missing",
+            context={"collector_name": datasource.source},
+        )
         return
 
     collector_registry.set_active(datasource.source, datasource.is_active)
@@ -48,13 +76,17 @@ async def _apply_datasource_schedule(datasource: DataSource, session) -> None:
             replace_existing=True,
             kwargs={"collector_name": datasource.source},
         )
-        logger.info(
-            "Scheduled collector: %s (every %sm)",
-            datasource.source,
-            datasource.frequency_minutes,
+        logger.info_event(
+            "Scheduled collector",
+            event="collector.schedule.updated",
+            context={"collector_name": datasource.source, "frequency_minutes": datasource.frequency_minutes},
         )
     else:
-        logger.info("Collector disabled: %s", datasource.source)
+        logger.info_event(
+            "Collector disabled",
+            event="collector.schedule.disabled",
+            context={"collector_name": datasource.source},
+        )
 
     await _update_next_run_at(datasource, session)
 
@@ -63,18 +95,30 @@ async def run_collector_task(collector_name: str):
     """Run a single collector task."""
     collector = collector_registry.get(collector_name)
     if not collector:
-        logger.error("Collector not found: %s", collector_name)
+        logger.error_event(
+            "Collector not found",
+            event="collector.run.collector_missing",
+            context={"collector_name": collector_name},
+        )
         return
 
     async with async_session_factory() as db:
         result = await db.execute(select(DataSource).where(DataSource.source == collector_name))
         datasource = result.scalar_one_or_none()
         if not datasource:
-            logger.error("Datasource not found for collector: %s", collector_name)
+            logger.error_event(
+                "Datasource not found for collector",
+                event="collector.run.datasource_missing",
+                context={"collector_name": collector_name},
+            )
             return
 
         if not datasource.is_active:
-            logger.info("Skipping disabled collector: %s", collector_name)
+            logger.info_event(
+                "Skipping disabled collector",
+                event="collector.run.skipped_disabled",
+                context={"collector_name": collector_name},
+            )
             return
 
         running_result = await db.execute(
@@ -98,10 +142,10 @@ async def run_collector_task(collector_name: str):
                 and (now - started_at) > timedelta(minutes=RUNNING_TASK_GUARD_TIMEOUT_MINUTES)
             )
             if not is_stale:
-                logger.warning(
-                    "Skipping collector %s trigger because task %s is already running",
-                    collector_name,
-                    existing_running.id,
+                logger.warning_event(
+                    "Skipping collector trigger because task is already running",
+                    event="collector.run.skipped_already_running",
+                    context={"collector_name": collector_name, "task_id": existing_running.id},
                 )
                 return
 
@@ -119,25 +163,47 @@ async def run_collector_task(collector_name: str):
                 else stale_reason
             )
             await db.commit()
-            logger.warning(
-                "Marked stale running task %s as failed before rerun of %s",
-                existing_running.id,
-                collector_name,
+            logger.warning_event(
+                "Marked stale running task as failed before rerun",
+                event="collector.run.stale_task_failed",
+                context={"collector_name": collector_name, "task_id": existing_running.id},
             )
 
         try:
             collector._datasource_id = datasource.id
-            logger.info("Running collector: %s (datasource_id=%s)", collector_name, datasource.id)
+            logger.info_event(
+                "Running collector",
+                event="collector.run.started",
+                context={"collector_name": collector_name, "datasource_id": datasource.id},
+            )
             task_result = await collector.run(db)
             datasource.last_run_at = datetime.now(UTC)
             datasource.last_status = task_result.get("status")
             await _update_next_run_at(datasource, db)
-            logger.info("Collector %s completed: %s", collector_name, task_result)
+            logger.info_event(
+                "Collector completed",
+                event="collector.run.completed",
+                context={"collector_name": collector_name, "datasource_id": datasource.id, "result": task_result},
+            )
+        except asyncio.CancelledError:
+            datasource.last_run_at = datetime.now(UTC)
+            datasource.last_status = "cancelled"
+            await db.commit()
+            logger.warning_event(
+                "Collector cancelled by operator",
+                event="collector.run.cancelled",
+                context={"collector_name": collector_name, "datasource_id": datasource.id},
+            )
+            raise
         except Exception as exc:
             datasource.last_run_at = datetime.now(UTC)
             datasource.last_status = "failed"
             await db.commit()
-            logger.exception("Collector %s failed: %s", collector_name, exc)
+            logger.exception_event(
+                "Collector failed",
+                event="collector.run.failed",
+                context={"collector_name": collector_name, "datasource_id": datasource.id, "error": str(exc)},
+            )
 
 
 async def cleanup_stale_running_tasks(max_age_hours: int = 2) -> int:
@@ -164,7 +230,11 @@ async def cleanup_stale_running_tasks(max_age_hours: int = 2) -> int:
 
         if stale_tasks:
             await db.commit()
-            logger.warning("Cleaned up %s stale running collection task(s)", len(stale_tasks))
+            logger.warning_event(
+                "Cleaned up stale running collection tasks",
+                event="collector.cleanup.stale_tasks_cleaned",
+                context={"count": len(stale_tasks)},
+            )
 
         return len(stale_tasks)
 
@@ -173,14 +243,14 @@ def start_scheduler() -> None:
     """Start the scheduler."""
     if not scheduler.running:
         scheduler.start()
-        logger.info("Scheduler started")
+        logger.info_event("Scheduler started", event="scheduler.started")
 
 
 def stop_scheduler() -> None:
     """Stop the scheduler."""
     if scheduler.running:
         scheduler.shutdown(wait=False)
-        logger.info("Scheduler stopped")
+        logger.info_event("Scheduler stopped", event="scheduler.stopped")
 
 
 async def sync_scheduler_with_datasources() -> None:
@@ -241,13 +311,56 @@ def run_collector_now(collector_name: str) -> bool:
     """Run a collector immediately (not scheduled)."""
     collector = collector_registry.get(collector_name)
     if not collector:
-        logger.error("Collector not found: %s", collector_name)
+        logger.error_event(
+            "Collector not found",
+            event="collector.trigger.collector_missing",
+            context={"collector_name": collector_name},
+        )
+        return False
+
+    existing_task = get_running_collector_task(collector_name)
+    if existing_task is not None and not existing_task.done():
+        logger.warning_event(
+            "Collector is already running in-memory; skipping duplicate trigger",
+            event="collector.trigger.skipped_already_running",
+            context={"collector_name": collector_name},
+        )
         return False
 
     try:
-        asyncio.create_task(run_collector_task(collector_name))
-        logger.info("Triggered collector: %s", collector_name)
+        task = asyncio.create_task(run_collector_task(collector_name), name=_collector_task_name(collector_name))
+        RUNNING_COLLECTOR_TASKS[collector_name] = task
+
+        def _cleanup_task(done_task: asyncio.Task[Any]) -> None:
+            current = RUNNING_COLLECTOR_TASKS.get(collector_name)
+            if current is done_task:
+                RUNNING_COLLECTOR_TASKS.pop(collector_name, None)
+
+        task.add_done_callback(_cleanup_task)
+        logger.info_event(
+            "Triggered collector",
+            event="collector.trigger.started",
+            context={"collector_name": collector_name},
+        )
         return True
     except Exception as exc:
-        logger.error("Failed to trigger collector %s: %s", collector_name, exc)
-        return False
+        logger.error_event(
+            "Failed to trigger collector",
+            event="collector.trigger.failed",
+            context={"collector_name": collector_name, "error": str(exc)},
+        )
+        return False
+
+
+async def cancel_running_collector_now(collector_name: str) -> bool:
+    task = get_running_collector_task(collector_name)
+    if task is None or task.done():
+        RUNNING_COLLECTOR_TASKS.pop(collector_name, None)
+        return False
+
+    task.cancel()
+    try:
+        await task
+    except asyncio.CancelledError:
+        return True
+    return task.cancelled()

backend/app/services/situational_alert_ai_brief.py

@@ -0,0 +1,174 @@
+from __future__ import annotations
+
+from collections import Counter
+from typing import Any
+
+from sqlalchemy import func, select
+from sqlalchemy.ext.asyncio import AsyncSession
+
+from app.models.alert import Alert, AlertSeverity, AlertStatus
+from app.models.bgp_anomaly import BGPAnomaly
+from app.models.bgp_incident import BGPIncident
+from app.schemas.ai import SituationalAnalysisRequest
+from app.services.bgp_ai_brief_store import get_latest_bgp_brief_record
+
+
+def _format_pairs(pairs: list[tuple[str, int]], empty_text: str = "无") -> str:
+    if not pairs:
+        return empty_text
+    return "，".join(f"{key} {value}" for key, value in pairs if key)
+
+
+async def build_situational_alert_brief_request(
+    db: AsyncSession,
+) -> tuple[SituationalAnalysisRequest, list[str], dict[str, Any]]:
+    total_alerts_result = await db.execute(select(func.count(Alert.id)))
+    active_alerts_result = await db.execute(
+        select(func.count(Alert.id)).where(Alert.status == AlertStatus.ACTIVE)
+    )
+    alert_severity_result = await db.execute(
+        select(Alert.severity, func.count(Alert.id))
+        .where(Alert.status == AlertStatus.ACTIVE)
+        .group_by(Alert.severity)
+    )
+    alert_source_result = await db.execute(
+        select(Alert.datasource_name, func.count(Alert.id))
+        .where(Alert.status == AlertStatus.ACTIVE)
+        .group_by(Alert.datasource_name)
+        .order_by(func.count(Alert.id).desc())
+        .limit(6)
+    )
+    recent_alerts_result = await db.execute(
+        select(Alert)
+        .order_by(Alert.created_at.desc(), Alert.id.desc())
+        .limit(6)
+    )
+
+    total_incidents_result = await db.execute(select(func.count(BGPIncident.id)))
+    active_incidents_result = await db.execute(
+        select(func.count(BGPIncident.id)).where(BGPIncident.status == "active")
+    )
+    bgp_severity_result = await db.execute(
+        select(BGPIncident.severity, func.count(BGPIncident.id))
+        .where(BGPIncident.status == "active")
+        .group_by(BGPIncident.severity)
+    )
+    bgp_region_counter: Counter[str] = Counter()
+    recent_incidents_result = await db.execute(
+        select(BGPIncident)
+        .order_by(BGPIncident.created_at.desc(), BGPIncident.id.desc())
+        .limit(5)
+    )
+
+    total_anomalies_result = await db.execute(select(func.count(BGPAnomaly.id)))
+    active_anomalies_result = await db.execute(
+        select(func.count(BGPAnomaly.id)).where(BGPAnomaly.status == "active")
+    )
+    anomaly_type_result = await db.execute(
+        select(BGPAnomaly.anomaly_type, func.count(BGPAnomaly.id))
+        .where(BGPAnomaly.status == "active")
+        .group_by(BGPAnomaly.anomaly_type)
+        .order_by(func.count(BGPAnomaly.id).desc())
+        .limit(6)
+    )
+
+    recent_incidents = recent_incidents_result.scalars().all()
+    for incident in recent_incidents:
+        for region in incident.affected_regions or []:
+            if not isinstance(region, dict):
+                continue
+            label = ", ".join(part for part in [region.get("city"), region.get("country")] if part) or "未知区域"
+            bgp_region_counter[label] += 1
+
+    latest_bgp_brief = get_latest_bgp_brief_record()
+    active_alert_severities = [
+        (item[0].value if isinstance(item[0], AlertSeverity) else str(item[0]), item[1])
+        for item in alert_severity_result.fetchall()
+        if item[0]
+    ]
+    active_bgp_severities = [
+        (str(item[0]), item[1])
+        for item in bgp_severity_result.fetchall()
+        if item[0]
+    ]
+    active_anomaly_types = [(str(item[0]), item[1]) for item in anomaly_type_result.fetchall() if item[0]]
+    active_alert_sources = [
+        (str(item[0] or "未命名数据源"), item[1])
+        for item in alert_source_result.fetchall()
+    ]
+
+    facts = [
+        (
+            f"系统告警侧：总告警 {total_alerts_result.scalar() or 0} 条，active {active_alerts_result.scalar() or 0} 条；"
+            f"活跃告警严重度分布为 {_format_pairs(active_alert_severities)}。"
+        ),
+        (
+            f"BGP态势侧：累计 incidents {total_incidents_result.scalar() or 0} 条，active incidents {active_incidents_result.scalar() or 0} 条；"
+            f"活跃 incidents 严重度分布为 {_format_pairs(active_bgp_severities)}。"
+        ),
+        (
+            f"BGP异常侧：累计 anomalies {total_anomalies_result.scalar() or 0} 条，active anomalies {active_anomalies_result.scalar() or 0} 条；"
+            f"活跃 anomaly 类型分布为 {_format_pairs(active_anomaly_types)}。"
+        ),
+    ]
+
+    if active_alert_sources:
+        facts.append(f"当前系统告警主要集中在：{_format_pairs(active_alert_sources)}。")
+    if bgp_region_counter:
+        facts.append(f"BGP近期高风险区域线索：{_format_pairs(bgp_region_counter.most_common(5))}。")
+
+    recent_alerts = recent_alerts_result.scalars().all()
+    if recent_alerts:
+        facts.append(
+            "最近系统告警摘录："
+            + "；".join(
+                [
+                    f"{alert.datasource_name or '未命名数据源'} / {alert.severity.value if alert.severity else '-'} / {alert.status.value if alert.status else '-'} / {alert.message or '-'}"
+                    for alert in recent_alerts
+                ]
+            )
+        )
+
+    if recent_incidents:
+        facts.append(
+            "最近BGP事件摘录："
+            + "；".join(
+                [
+                    f"{incident.incident_type} / {incident.severity} / {incident.status} / {incident.summary}"
+                    for incident in recent_incidents
+                ]
+            )
+        )
+
+    if latest_bgp_brief:
+        facts.append(
+            f"最近一份 BGP AI 简报生成于 {latest_bgp_brief.generated_at}，模型 {latest_bgp_brief.model}，可作为当前态势的补充说明。"
+        )
+
+    context = {
+        "source": "situational-alerts",
+        "active_system_alerts": active_alerts_result.scalar() or 0,
+        "active_system_alert_severities": dict(active_alert_severities),
+        "top_system_alert_sources": dict(active_alert_sources),
+        "active_bgp_incidents": active_incidents_result.scalar() or 0,
+        "active_bgp_incident_severities": dict(active_bgp_severities),
+        "active_bgp_anomalies": active_anomalies_result.scalar() or 0,
+        "active_bgp_anomaly_types": dict(active_anomaly_types),
+        "bgp_hot_regions": dict(bgp_region_counter.most_common(5)),
+        "latest_bgp_brief_id": latest_bgp_brief.id if latest_bgp_brief else None,
+        "latest_bgp_brief_generated_at": latest_bgp_brief.generated_at if latest_bgp_brief else None,
+    }
+
+    request = SituationalAnalysisRequest(
+        title="态势告警 AI 简报",
+        objective="综合系统告警、BGP incidents、BGP anomalies 与近期 BGP AI 简报，生成一份面向值班人员的态势告警简报，指出当前最需要关注的风险域、跨模块联动迹象和优先动作。",
+        observations=facts,
+        constraints=[
+            "明确区分事实、推断与建议。",
+            "优先指出仍在 active 状态的系统告警与 BGP 风险是否存在联动。",
+            "不要把单一数据源的局部异常夸大成全局态势。",
+            "如果证据不足，请明确写出仍缺哪些模块或区域信息。",
+        ],
+        context=context,
+    )
+    return request, facts, context

backend/app/services/system_logs.py

@@ -0,0 +1,532 @@
+from __future__ import annotations
+
+import json
+import re
+import shutil
+import subprocess
+
+from collections import Counter, deque
+from dataclasses import dataclass
+from datetime import UTC, datetime
+from pathlib import Path
+from typing import Any
+
+from app.core.security import redis_client
+
+DEFAULT_LOG_LINE_LIMIT = 200
+MAX_LOG_LINE_LIMIT = 1000
+BUFFER_LOG_LIMIT = 1000
+BUFFER_LOG_TTL_SECONDS = 7 * 24 * 60 * 60
+LOG_BUFFER_KEY_PREFIX = "planet:system_logs"
+
+LOG_LEVEL_ERROR = "error"
+LOG_LEVEL_WARNING = "warning"
+LOG_LEVEL_INFO = "info"
+LOG_LEVEL_DEBUG = "debug"
+LOG_LEVEL_ALL = "all"
+
+SUPPORTED_LOG_LEVELS = {
+    LOG_LEVEL_ALL,
+    LOG_LEVEL_ERROR,
+    LOG_LEVEL_WARNING,
+    LOG_LEVEL_INFO,
+    LOG_LEVEL_DEBUG,
+}
+
+LOG_LEVEL_ALIASES = {
+    "warn": LOG_LEVEL_WARNING,
+    "warning": LOG_LEVEL_WARNING,
+    "err": LOG_LEVEL_ERROR,
+    "error": LOG_LEVEL_ERROR,
+    "info": LOG_LEVEL_INFO,
+    "information": LOG_LEVEL_INFO,
+    "debug": LOG_LEVEL_DEBUG,
+    "trace": LOG_LEVEL_DEBUG,
+    "critical": LOG_LEVEL_ERROR,
+    "fatal": LOG_LEVEL_ERROR,
+}
+
+TIMESTAMP_FORMATS = (
+    "%Y-%m-%d %H:%M:%S.%f",
+    "%Y-%m-%d %H:%M:%S",
+    "%Y-%m-%dT%H:%M:%S.%f",
+    "%Y-%m-%dT%H:%M:%S",
+)
+
+LEVEL_PATTERNS = (
+    ("CRITICAL", LOG_LEVEL_ERROR),
+    ("FATAL", LOG_LEVEL_ERROR),
+    ("ERROR", LOG_LEVEL_ERROR),
+    ("WARNING", LOG_LEVEL_WARNING),
+    ("WARN", LOG_LEVEL_WARNING),
+    ("INFO", LOG_LEVEL_INFO),
+    ("DEBUG", LOG_LEVEL_DEBUG),
+    ("TRACE", LOG_LEVEL_DEBUG),
+)
+
+LEADING_LEVEL_PATTERN = re.compile(
+    r"^\s*(?:\[[^\]]+\]\s*)?(CRITICAL|FATAL|ERROR|WARNING|WARN|INFO|DEBUG|TRACE)\b[:\s-]*",
+    re.IGNORECASE,
+)
+EMBEDDED_LEVEL_PATTERN = re.compile(
+    r"\b(CRITICAL|FATAL|ERROR|WARNING|WARN|INFO|DEBUG|TRACE)\b",
+    re.IGNORECASE,
+)
+CONTROL_CHAR_PATTERN = re.compile(r"[\x00-\x08\x0b\x0c\x0e-\x1f\x7f]")
+
+
+@dataclass(frozen=True)
+class LogSource:
+    source_id: str
+    name: str
+    kind: str
+    location: str
+    description: str
+    category: str
+    status: str = "ok"
+    buffer_key: str | None = None
+    container_name: str | None = None
+
+
+@dataclass
+class StructuredLogEntry:
+    timestamp: datetime | None
+    level: str | None
+    display_line: str
+    raw_line: str
+    search_text: str
+
+
+@dataclass
+class DailyLogMarker:
+    date_token: str
+    total: int
+    dominant_level: str
+
+
+LOG_SOURCES: dict[str, LogSource] = {
+    "backend": LogSource(
+        source_id="backend",
+        name="后端服务",
+        kind="file",
+        location="/tmp/planet_backend.log",
+        description="FastAPI 后端、调度器和采集任务共享日志。",
+        category="service",
+    ),
+    "frontend": LogSource(
+        source_id="frontend",
+        name="前端开发服务",
+        kind="file",
+        location="/tmp/planet_frontend.log",
+        description="控制台与 Earth 前端开发服务输出。",
+        category="service",
+    ),
+    "ai-provider": LogSource(
+        source_id="ai-provider",
+        name="AI Provider",
+        kind="docker",
+        location="docker://planet_aiprovider",
+        description="AI Provider 容器实时输出日志。",
+        category="service",
+        container_name="planet_aiprovider",
+    ),
+    "earth-client": LogSource(
+        source_id="earth-client",
+        name="Earth 浏览器端",
+        kind="buffer",
+        location="redis://planet:system_logs:earth-client",
+        description="Earth 浏览器端上报的运行时错误与关键业务日志。",
+        category="client",
+        buffer_key=f"{LOG_BUFFER_KEY_PREFIX}:earth-client",
+    ),
+}
+
+
+def normalize_log_level(level: str | None) -> str:
+    if level is None:
+        return LOG_LEVEL_ALL
+    normalized = str(level).strip().lower()
+    if normalized in {"", LOG_LEVEL_ALL}:
+        return LOG_LEVEL_ALL
+    return LOG_LEVEL_ALIASES.get(normalized, LOG_LEVEL_ALL)
+
+
+def normalize_log_levels(level: str | None = None, levels: str | None = None) -> tuple[str, ...]:
+    normalized_levels: list[str] = []
+    if levels:
+        for item in str(levels).split(","):
+            normalized = normalize_log_level(item)
+            if normalized != LOG_LEVEL_ALL and normalized not in normalized_levels:
+                normalized_levels.append(normalized)
+    normalized_level = normalize_log_level(level)
+    if normalized_level != LOG_LEVEL_ALL and normalized_level not in normalized_levels:
+        normalized_levels.append(normalized_level)
+    return tuple(normalized_levels)
+
+
+def get_source_status(source: LogSource) -> str:
+    if source.kind == "file":
+        path = Path(source.location)
+        if not path.exists():
+            return "missing"
+        return "ok" if path.stat().st_size > 0 else "empty"
+    if source.kind == "docker":
+        return "ok" if shutil.which("docker") else "docker_unavailable"
+    if source.kind == "buffer":
+        if not source.buffer_key:
+            return "source_unavailable"
+        try:
+            return "ok" if redis_client.llen(source.buffer_key) > 0 else "empty"
+        except Exception:
+            return "source_unavailable"
+    return "source_unavailable"
+
+
+def list_log_sources() -> list[dict[str, str]]:
+    items: list[dict[str, str]] = []
+    for source in LOG_SOURCES.values():
+        items.append(
+            {
+                "source_id": source.source_id,
+                "name": source.name,
+                "kind": source.kind,
+                "location": source.location,
+                "description": source.description,
+                "category": source.category,
+                "status": get_source_status(source),
+            }
+        )
+    return items
+
+
+def get_buffer_log_key(source_id: str) -> str:
+    return f"{LOG_BUFFER_KEY_PREFIX}:{source_id}"
+
+
+def append_buffer_log(
+    source_id: str,
+    *,
+    level: str,
+    message: str,
+    context: dict[str, Any] | None = None,
+) -> None:
+    payload = {
+        "timestamp": datetime.now(tz=UTC).isoformat(),
+        "level": normalize_log_level(level),
+        "message": message,
+        "context": context or {},
+    }
+    buffer_key = get_buffer_log_key(source_id)
+    redis_client.rpush(buffer_key, json.dumps(payload, ensure_ascii=False))
+    redis_client.ltrim(buffer_key, -BUFFER_LOG_LIMIT, -1)
+    redis_client.expire(buffer_key, BUFFER_LOG_TTL_SECONDS)
+
+
+def parse_timestamp(raw_value: str | None) -> datetime | None:
+    if not raw_value:
+        return None
+    candidate = str(raw_value).strip()
+    if not candidate:
+        return None
+    candidate = candidate.replace("Z", "+00:00")
+    try:
+        parsed = datetime.fromisoformat(candidate)
+        return parsed if parsed.tzinfo else parsed.replace(tzinfo=UTC)
+    except ValueError:
+        pass
+    for fmt in TIMESTAMP_FORMATS:
+        try:
+            return datetime.strptime(candidate, fmt).replace(tzinfo=UTC)
+        except ValueError:
+            continue
+    return None
+
+
+def parse_prefixed_timestamp(line: str) -> tuple[datetime | None, str]:
+    stripped = line.strip()
+    if not stripped:
+        return None, ""
+    for prefix_length in (35, 32, 29, 26, 23, 19):
+        if len(stripped) < prefix_length:
+            continue
+        prefix = stripped[:prefix_length]
+        timestamp = parse_timestamp(prefix)
+        if timestamp is not None:
+            return timestamp, stripped[prefix_length:].lstrip()
+    first_token = stripped.split(maxsplit=1)[0]
+    timestamp = parse_timestamp(first_token)
+    if timestamp is not None:
+        remainder = stripped[len(first_token):].lstrip()
+        return timestamp, remainder
+    return None, stripped
+
+
+def infer_log_level_from_text(text: str, *, allow_embedded: bool = True) -> str | None:
+    leading_match = LEADING_LEVEL_PATTERN.match(text)
+    if leading_match:
+        return normalize_log_level(leading_match.group(1))
+
+    if allow_embedded:
+        embedded_match = EMBEDDED_LEVEL_PATTERN.search(text)
+        if embedded_match:
+            return normalize_log_level(embedded_match.group(1))
+        upper_text = text.upper()
+        for pattern, normalized in LEVEL_PATTERNS:
+            if f"{pattern}:" in upper_text or f"{pattern} " in upper_text:
+                return normalized
+    return None
+
+
+def build_display_line(timestamp: datetime | None, level: str | None, message: str) -> str:
+    message_part = message.strip() if message else ""
+    parts = []
+    if timestamp is not None:
+        parts.append(timestamp.astimezone(UTC).strftime("%Y-%m-%d %H:%M:%S"))
+    if level:
+        parts.append(level.upper())
+    if message_part:
+        parts.append(message_part)
+    return " ".join(parts).strip()
+
+
+def sanitize_text_log_line(line: str) -> str:
+    return CONTROL_CHAR_PATTERN.sub("", line)
+
+
+def parse_text_log_entry(line: str) -> StructuredLogEntry:
+    sanitized_line = sanitize_text_log_line(line).rstrip("\n")
+    timestamp, remainder = parse_prefixed_timestamp(sanitized_line)
+    level = infer_log_level_from_text(remainder or sanitized_line, allow_embedded=False)
+    display_line = sanitized_line
+    return StructuredLogEntry(
+        timestamp=timestamp,
+        level=level,
+        display_line=display_line,
+        raw_line=display_line,
+        search_text=display_line.lower(),
+    )
+
+
+def build_buffer_entry(payload: dict[str, Any]) -> StructuredLogEntry:
+    timestamp = parse_timestamp(str(payload.get("timestamp", "")).strip())
+    level = normalize_log_level(payload.get("level"))
+    if level == LOG_LEVEL_ALL:
+        level = None
+    message = str(payload.get("message", "")).strip()
+    context = payload.get("context")
+    context_map = context if isinstance(context, dict) else {}
+    context_fragments = []
+    for key in ("category", "module", "url", "detail"):
+        value = str(context_map.get(key, "")).strip()
+        if value:
+            context_fragments.append(f"{key}={value}")
+    message_with_context = " | ".join([message, *context_fragments]) if context_fragments else message
+    display_line = build_display_line(timestamp, level, message_with_context)
+    search_text = " ".join(
+        [
+            message,
+            json.dumps(context_map, ensure_ascii=False, sort_keys=True),
+            display_line,
+        ]
+    ).lower()
+    return StructuredLogEntry(
+        timestamp=timestamp,
+        level=level,
+        display_line=display_line,
+        raw_line=json.dumps(payload, ensure_ascii=False, sort_keys=True),
+        search_text=search_text,
+    )
+
+
+def read_file_entries(source: LogSource, scan_limit: int) -> list[StructuredLogEntry]:
+    path = Path(source.location)
+    if not path.exists():
+        return []
+    with path.open("r", encoding="utf-8", errors="replace") as handle:
+        recent_lines = deque(handle, maxlen=scan_limit)
+    return [
+        parse_text_log_entry(line)
+        for line in recent_lines
+        if sanitize_text_log_line(line).strip()
+    ]
+
+
+def read_docker_entries(source: LogSource, scan_limit: int) -> list[StructuredLogEntry]:
+    if not shutil.which("docker") or not source.container_name:
+        return []
+    try:
+        completed = subprocess.run(
+            [
+                "docker",
+                "logs",
+                "--timestamps",
+                "--tail",
+                str(scan_limit),
+                source.container_name,
+            ],
+            capture_output=True,
+            text=True,
+            check=False,
+        )
+    except OSError:
+        return []
+    if completed.returncode != 0:
+        return []
+    return [
+        parse_text_log_entry(line)
+        for line in completed.stdout.splitlines()
+        if line.strip()
+    ]
+
+
+def read_buffer_entries(source: LogSource, scan_limit: int) -> list[StructuredLogEntry]:
+    if not source.buffer_key:
+        return []
+    try:
+        raw_items = redis_client.lrange(source.buffer_key, -scan_limit, -1)
+    except Exception:
+        return []
+    entries: list[StructuredLogEntry] = []
+    for raw_item in raw_items:
+        try:
+            payload = json.loads(raw_item)
+        except json.JSONDecodeError:
+            entries.append(parse_text_log_entry(str(raw_item)))
+            continue
+        if isinstance(payload, dict):
+            entries.append(build_buffer_entry(payload))
+        else:
+            entries.append(parse_text_log_entry(str(raw_item)))
+    return entries
+
+
+def read_source_entries(source: LogSource, scan_limit: int) -> list[StructuredLogEntry]:
+    if source.kind == "file":
+        return read_file_entries(source, scan_limit)
+    if source.kind == "docker":
+        return read_docker_entries(source, scan_limit)
+    if source.kind == "buffer":
+        return read_buffer_entries(source, scan_limit)
+    return []
+
+
+def matches_levels(entry: StructuredLogEntry, selected_levels: tuple[str, ...]) -> bool:
+    if not selected_levels:
+        return True
+    return entry.level in selected_levels
+
+
+def matches_date_range(
+    entry: StructuredLogEntry,
+    start_date: str | None,
+    end_date: str | None,
+) -> bool:
+    if not start_date and not end_date:
+        return True
+    if entry.timestamp is None:
+        return False
+    date_token = entry.timestamp.astimezone(UTC).date().isoformat()
+    if start_date and date_token < start_date:
+        return False
+    if end_date and date_token > end_date:
+        return False
+    return True
+
+
+def matches_search(entry: StructuredLogEntry, search: str | None) -> bool:
+    if search is None:
+        return True
+    query = search.strip().lower()
+    if not query:
+        return True
+    return query in entry.search_text
+
+
+def build_daily_log_markers(entries: list[StructuredLogEntry]) -> list[dict[str, Any]]:
+    grouped: dict[str, list[StructuredLogEntry]] = {}
+    for entry in entries:
+        if entry.timestamp is None:
+            continue
+        date_token = entry.timestamp.astimezone(UTC).date().isoformat()
+        grouped.setdefault(date_token, []).append(entry)
+
+    markers: list[DailyLogMarker] = []
+    for date_token, group in sorted(grouped.items()):
+        level_counts = Counter(
+            entry.level
+            for entry in group
+            if entry.level in SUPPORTED_LOG_LEVELS and entry.level != LOG_LEVEL_ALL
+        )
+        dominant_level = LOG_LEVEL_INFO
+        if level_counts:
+            dominant_level = sorted(
+                level_counts.items(),
+                key=lambda item: (
+                    -item[1],
+                    ("error", "warning", "info", "debug").index(item[0]),
+                ),
+            )[0][0]
+        markers.append(
+            DailyLogMarker(
+                date_token=date_token,
+                total=len(group),
+                dominant_level=dominant_level,
+            )
+        )
+    return [marker.__dict__ for marker in markers]
+
+
+def read_log_snapshot(
+    source_id: str,
+    limit: int,
+    *,
+    level: str = LOG_LEVEL_ALL,
+    levels: str | None = None,
+    start_date: str | None = None,
+    end_date: str | None = None,
+    search: str | None = None,
+) -> dict[str, Any] | None:
+    source = LOG_SOURCES.get(source_id)
+    if source is None:
+        return None
+
+    selected_levels = normalize_log_levels(level, levels)
+    search_query = (search or "").strip()
+    scan_limit = max(min(MAX_LOG_LINE_LIMIT * 5, 5000), limit * 5, BUFFER_LOG_LIMIT if source.kind == "buffer" else 1000)
+    all_entries = read_source_entries(source, scan_limit)
+    marker_entries = [
+        entry
+        for entry in all_entries
+        if matches_levels(entry, selected_levels) and matches_search(entry, search_query)
+    ]
+    filtered_entries = [
+        entry
+        for entry in marker_entries
+        if matches_date_range(entry, start_date, end_date)
+    ]
+    visible_entries = filtered_entries[-limit:]
+
+    compatibility_level = selected_levels[0] if len(selected_levels) == 1 else LOG_LEVEL_ALL
+    return {
+        "source_id": source.source_id,
+        "name": source.name,
+        "kind": source.kind,
+        "location": source.location,
+        "description": source.description,
+        "category": source.category,
+        "status": get_source_status(source),
+        "level": compatibility_level,
+        "selected_levels": list(selected_levels),
+        "search_query": search_query,
+        "available_levels": [
+            LOG_LEVEL_ALL,
+            LOG_LEVEL_ERROR,
+            LOG_LEVEL_WARNING,
+            LOG_LEVEL_INFO,
+            LOG_LEVEL_DEBUG,
+        ],
+        "daily_markers": build_daily_log_markers(marker_entries),
+        "line_limit": limit,
+        "line_count": len(visible_entries),
+        "lines": [entry.display_line for entry in visible_entries],
+    }

backend/app/services/tv_streams.py

@@ -0,0 +1,466 @@
+from __future__ import annotations
+
+from datetime import UTC, datetime
+from typing import Any
+from urllib.parse import urlparse
+
+from sqlalchemy import select
+from sqlalchemy.ext.asyncio import AsyncSession
+
+from app.core.time import to_iso8601_utc
+from app.models.collected_data import CollectedData
+from app.models.system_setting import SystemSetting
+
+DEFAULT_TV_SOURCE_ID = "cgtn-en"
+TV_SETTINGS_CATEGORY = "tv"
+TV_LIVE_SOURCE_COLLECTOR = "news_live_streams"
+TV_LIVE_SOURCE_DATA_TYPE = "news_live_stream"
+
+DEFAULT_TV_SETTINGS = {
+    "default_source_id": DEFAULT_TV_SOURCE_ID,
+    "auto_fallback": True,
+    "sources": [
+        {
+            "id": "cctv4",
+            "name": "CCTV-4 中文国际",
+            "provider": "CCTV",
+            "region": "China",
+            "language": "zh-CN",
+            "source_type": "hls",
+            "embed_url": "https://tv.cctv.com/live/cctv4/",
+            "stream_url": "https://ldocctvwbcdtxy.liveplay.myqcloud.com/ldocctvwbcd/cdrmldcctv4_1_td.m3u8",
+            "homepage_url": "https://tv.cctv.com/live/cctv4/",
+            "poster_url": "",
+            "is_enabled": True,
+            "is_fallback": True,
+            "sort_order": 10,
+            "collector_source": None,
+            "notes": "默认兜底新闻直播源。优先尝试 CCTV-4 官方 HLS 播放流，若直播放失败则回退到央视官网直播页。",
+        },
+        {
+            "id": "reuters-tv",
+            "name": "Reuters TV",
+            "provider": "Reuters",
+            "region": "Global",
+            "language": "en",
+            "source_type": "hls",
+            "embed_url": "",
+            "stream_url": "https://reuters-reutersnow-1-eu.rakuten.wurl.tv/playlist.m3u8",
+            "homepage_url": "https://www.reuters.com/video/live/",
+            "poster_url": "",
+            "is_enabled": True,
+            "is_fallback": False,
+            "sort_order": 20,
+            "collector_source": None,
+            "notes": "参考 worldmonitor 的默认新闻频道清单，优先作为全球英文新闻直播放源。",
+        },
+        {
+            "id": "cgtn-en",
+            "name": "CGTN English",
+            "provider": "CGTN",
+            "region": "Global",
+            "language": "en",
+            "source_type": "youtube",
+            "embed_url": "https://www.youtube.com/watch?v=BOy2xDU1LC8",
+            "stream_url": "https://news.cgtn.com/resource/live/english/cgtn-news.m3u8",
+            "youtube_video_id": "BOy2xDU1LC8",
+            "homepage_url": "https://news.cgtn.com/",
+            "poster_url": "",
+            "is_enabled": True,
+            "is_fallback": False,
+            "sort_order": 30,
+            "collector_source": None,
+            "notes": "优先使用官方 YouTube 直播源，保留 HLS 直播放流作为候选信息。",
+        },
+        {
+            "id": "cgtn-es",
+            "name": "CGTN Espanol",
+            "provider": "CGTN",
+            "region": "Latin America",
+            "language": "es",
+            "source_type": "hls",
+            "embed_url": "",
+            "stream_url": "https://news.cgtn.com/resource/live/espanol/cgtn-e.m3u8",
+            "homepage_url": "https://news.cgtn.com/",
+            "poster_url": "",
+            "is_enabled": True,
+            "is_fallback": False,
+            "sort_order": 40,
+            "collector_source": None,
+            "notes": "西语国际新闻频道，覆盖拉美方向态势。",
+        },
+        {
+            "id": "dw-espanol",
+            "name": "DW Espanol",
+            "provider": "Deutsche Welle",
+            "region": "Europe",
+            "language": "es",
+            "source_type": "hls",
+            "embed_url": "",
+            "stream_url": "https://dwamdstream104.akamaized.net/hls/live/2015530/dwstream104/stream04/streamPlaylist.m3u8",
+            "homepage_url": "https://www.dw.com/es/",
+            "poster_url": "",
+            "is_enabled": True,
+            "is_fallback": False,
+            "sort_order": 50,
+            "collector_source": None,
+            "notes": "来自 worldmonitor 可选频道清单的直播放源。",
+        },
+        {
+            "id": "dw-arabic",
+            "name": "DW Arabic",
+            "provider": "Deutsche Welle",
+            "region": "Middle East",
+            "language": "ar",
+            "source_type": "hls",
+            "embed_url": "",
+            "stream_url": "https://dwamdstream103.akamaized.net/hls/live/2015526/dwstream103/index.m3u8",
+            "homepage_url": "https://www.dw.com/ar/",
+            "poster_url": "",
+            "is_enabled": True,
+            "is_fallback": False,
+            "sort_order": 60,
+            "collector_source": None,
+            "notes": "阿拉伯语新闻流，适合作为中东方向新闻补充源。",
+        },
+        {
+            "id": "aljazeera-mubasher",
+            "name": "Al Jazeera Mubasher",
+            "provider": "Al Jazeera",
+            "region": "Middle East",
+            "language": "ar",
+            "source_type": "hls",
+            "embed_url": "",
+            "stream_url": "https://live-hls-web-ajm.getaj.net/AJM/index.m3u8",
+            "homepage_url": "https://www.aljazeera.net/live",
+            "poster_url": "",
+            "is_enabled": True,
+            "is_fallback": False,
+            "sort_order": 70,
+            "collector_source": None,
+            "notes": "中东实时新闻流，来自 worldmonitor HLS 频道目录。",
+        },
+        {
+            "id": "arirang-news",
+            "name": "Arirang News",
+            "provider": "Arirang",
+            "region": "Korea",
+            "language": "en",
+            "source_type": "hls",
+            "embed_url": "",
+            "stream_url": "https://amdlive-ch01-ctnd-com.akamaized.net/arirang_1ch/smil:arirang_1ch.smil/playlist.m3u8",
+            "homepage_url": "https://www.arirang.com/",
+            "poster_url": "",
+            "is_enabled": True,
+            "is_fallback": False,
+            "sort_order": 80,
+            "collector_source": None,
+            "notes": "东北亚英语新闻源，适合补充韩半岛与东亚视角。",
+        },
+        {
+            "id": "abp-news",
+            "name": "ABP News",
+            "provider": "ABP",
+            "region": "India",
+            "language": "hi",
+            "source_type": "hls",
+            "embed_url": "",
+            "stream_url": "https://abplivetv.pc.cdn.bitgravity.com/httppush/abp_livetv/abp_abpnews/master.m3u8",
+            "homepage_url": "https://news.abplive.com/live-tv",
+            "poster_url": "",
+            "is_enabled": True,
+            "is_fallback": False,
+            "sort_order": 90,
+            "collector_source": None,
+            "notes": "印度新闻直播放源，补充南亚区域视角。",
+        },
+        {
+            "id": "sabc-news",
+            "name": "SABC News",
+            "provider": "SABC",
+            "region": "Africa",
+            "language": "en",
+            "source_type": "hls",
+            "embed_url": "",
+            "stream_url": "https://sabconetanw.cdn.mangomolo.com/news/smil:news.stream.smil/playlist.m3u8",
+            "homepage_url": "https://www.sabcnews.com/sabcnews/",
+            "poster_url": "",
+            "is_enabled": True,
+            "is_fallback": False,
+            "sort_order": 100,
+            "collector_source": None,
+            "notes": "非洲英语新闻源，补充非洲区域新闻覆盖。",
+        },
+    ],
+}
+
+
+def _clean_text(value: Any) -> str:
+    if value is None:
+        return ""
+    return str(value).strip()
+
+
+def _clean_url(value: Any) -> str:
+    text = _clean_text(value)
+    if not text:
+        return ""
+
+    parsed = urlparse(text)
+    if parsed.scheme and parsed.scheme not in {"http", "https"}:
+        return ""
+    if parsed.scheme and not parsed.netloc:
+        return ""
+    return text
+
+
+def _clean_bool(value: Any, *, default: bool) -> bool:
+    if isinstance(value, bool):
+        return value
+    if value in (None, ""):
+        return default
+    if isinstance(value, str):
+        lowered = value.strip().lower()
+        if lowered in {"1", "true", "yes", "on"}:
+            return True
+        if lowered in {"0", "false", "no", "off"}:
+            return False
+    return bool(value)
+
+
+def _clean_int(value: Any, *, default: int) -> int:
+    try:
+        return int(value)
+    except (TypeError, ValueError):
+        return default
+
+
+def normalize_tv_source(source: dict[str, Any] | None, *, index: int = 0) -> dict[str, Any]:
+    payload = dict(source or {})
+    source_id = _clean_text(payload.get("id")) or f"tv-source-{index + 1}"
+    source_type = _clean_text(payload.get("source_type")).lower()
+    youtube_video_id = _clean_text(payload.get("youtube_video_id"))
+    youtube_channel = _clean_text(payload.get("youtube_channel"))
+    if source_type not in {"iframe", "hls", "video", "external", "youtube"}:
+        if youtube_video_id or youtube_channel:
+            source_type = "youtube"
+        else:
+            source_type = "iframe" if _clean_text(payload.get("embed_url")) else "external"
+
+    if source_type == "youtube" and not youtube_video_id and not youtube_channel:
+        source_type = "iframe" if _clean_text(payload.get("embed_url")) else "external"
+
+    return {
+        "id": source_id,
+        "name": _clean_text(payload.get("name")) or f"新闻直播源 {index + 1}",
+        "provider": _clean_text(payload.get("provider")) or "Unknown",
+        "region": _clean_text(payload.get("region")) or "Global",
+        "language": _clean_text(payload.get("language")) or "und",
+        "source_type": source_type,
+        "embed_url": _clean_url(payload.get("embed_url")),
+        "stream_url": _clean_url(payload.get("stream_url")),
+        "homepage_url": _clean_url(payload.get("homepage_url")),
+        "poster_url": _clean_url(payload.get("poster_url")),
+        "youtube_video_id": youtube_video_id,
+        "youtube_channel": youtube_channel,
+        "is_enabled": _clean_bool(payload.get("is_enabled"), default=True),
+        "is_fallback": _clean_bool(payload.get("is_fallback"), default=False),
+        "sort_order": _clean_int(payload.get("sort_order"), default=(index + 1) * 10),
+        "collector_source": payload.get("collector_source"),
+        "notes": _clean_text(payload.get("notes")),
+        "updated_at": _clean_text(payload.get("updated_at")),
+    }
+
+
+def normalize_tv_settings(payload: dict[str, Any] | None) -> dict[str, Any]:
+    merged = {
+        "default_source_id": DEFAULT_TV_SETTINGS["default_source_id"],
+        "auto_fallback": DEFAULT_TV_SETTINGS["auto_fallback"],
+        "sources": [],
+    }
+
+    raw_sources = []
+    if isinstance(payload, dict):
+        merged["default_source_id"] = (
+            _clean_text(payload.get("default_source_id")) or merged["default_source_id"]
+        )
+        merged["auto_fallback"] = _clean_bool(
+            payload.get("auto_fallback"),
+            default=DEFAULT_TV_SETTINGS["auto_fallback"],
+        )
+        if isinstance(payload.get("sources"), list):
+            raw_sources = payload["sources"]
+
+    if not raw_sources:
+        raw_sources = DEFAULT_TV_SETTINGS["sources"]
+
+    normalized_sources = [
+        normalize_tv_source(source, index=index)
+        for index, source in enumerate(raw_sources)
+    ]
+
+    if not any(source["id"] == DEFAULT_TV_SOURCE_ID for source in normalized_sources):
+        normalized_sources.append(
+            normalize_tv_source(DEFAULT_TV_SETTINGS["sources"][0], index=len(normalized_sources))
+        )
+
+    default_source_exists = any(
+        source["id"] == merged["default_source_id"] and source["is_enabled"]
+        for source in normalized_sources
+    )
+    if not default_source_exists:
+        fallback_source = next(
+            (source for source in normalized_sources if source["is_fallback"] and source["is_enabled"]),
+            None,
+        )
+        first_enabled_source = next(
+            (source for source in normalized_sources if source["is_enabled"]),
+            None,
+        )
+        merged["default_source_id"] = (
+            fallback_source["id"]
+            if fallback_source
+            else first_enabled_source["id"]
+            if first_enabled_source
+            else DEFAULT_TV_SOURCE_ID
+        )
+
+    merged["sources"] = sorted(
+        normalized_sources,
+        key=lambda item: (item["sort_order"], item["name"], item["id"]),
+    )
+    return merged
+
+
+async def get_tv_settings_payload(db: AsyncSession) -> dict[str, Any]:
+    result = await db.execute(
+        select(SystemSetting).where(SystemSetting.category == TV_SETTINGS_CATEGORY)
+    )
+    record = result.scalar_one_or_none()
+    payload = record.payload if record else None
+    return normalize_tv_settings(payload)
+
+
+def _build_collected_tv_source(record: CollectedData, index: int) -> dict[str, Any]:
+    metadata = dict(record.extra_data or {})
+    return normalize_tv_source(
+        {
+            "id": metadata.get("id") or record.source_id or record.entity_key,
+            "name": record.name or record.title or metadata.get("name") or f"采集直播源 {index + 1}",
+            "provider": metadata.get("provider") or metadata.get("publisher") or "Collector",
+            "region": metadata.get("region") or metadata.get("country") or "Global",
+            "language": metadata.get("language") or "und",
+            "source_type": metadata.get("source_type") or "iframe",
+            "embed_url": metadata.get("embed_url") or metadata.get("url") or "",
+            "stream_url": metadata.get("stream_url") or "",
+            "homepage_url": metadata.get("homepage_url") or metadata.get("source_url") or "",
+            "poster_url": metadata.get("poster_url") or "",
+            "youtube_video_id": metadata.get("youtube_video_id") or metadata.get("video_id") or "",
+            "youtube_channel": metadata.get("youtube_channel") or metadata.get("channel_handle") or "",
+            "is_enabled": metadata.get("is_enabled", True),
+            "is_fallback": False,
+            "sort_order": metadata.get("sort_order", 200 + index),
+            "collector_source": record.source,
+            "notes": record.description or metadata.get("notes") or "",
+            "updated_at": to_iso8601_utc(record.collected_at or record.reference_date or datetime.now(UTC)),
+        },
+        index=index,
+    )
+
+
+async def get_collected_tv_sources(db: AsyncSession) -> list[dict[str, Any]]:
+    result = await db.execute(
+        select(CollectedData)
+        .where(CollectedData.source == TV_LIVE_SOURCE_COLLECTOR)
+        .where(CollectedData.data_type == TV_LIVE_SOURCE_DATA_TYPE)
+        .where(CollectedData.is_current.is_(True))
+        .where(CollectedData.is_valid == 1)
+        .order_by(CollectedData.reference_date.desc().nullslast(), CollectedData.id.desc())
+    )
+    rows = result.scalars().all()
+    return [_build_collected_tv_source(record, index) for index, record in enumerate(rows)]
+
+
+def build_public_tv_payload(
+    settings_payload: dict[str, Any],
+    collected_sources: list[dict[str, Any]],
+) -> dict[str, Any]:
+    configured_sources = [
+        source for source in settings_payload["sources"] if source["is_enabled"]
+    ]
+
+    merged_by_id = {source["id"]: source for source in configured_sources}
+    for source in collected_sources:
+        if source["id"] in merged_by_id or not source["is_enabled"]:
+            continue
+        merged_by_id[source["id"]] = source
+
+    available_sources = sorted(
+        merged_by_id.values(),
+        key=lambda item: (item["sort_order"], item["name"], item["id"]),
+    )
+
+    default_source = next(
+        (
+            source
+            for source in available_sources
+            if source["id"] == settings_payload["default_source_id"]
+        ),
+        None,
+    )
+    fallback_source = next(
+        (source for source in available_sources if source["is_fallback"]),
+        None,
+    )
+
+    resolved_source = default_source or fallback_source or (available_sources[0] if available_sources else None)
+    latest_updated_at = max(
+        (source.get("updated_at") or "" for source in available_sources),
+        default="",
+    )
+
+    return {
+        "default_source_id": settings_payload["default_source_id"],
+        "auto_fallback": settings_payload["auto_fallback"],
+        "selected_source": resolved_source,
+        "fallback_source": fallback_source,
+        "sources": available_sources,
+        "source_count": len(available_sources),
+        "latest_updated_at": latest_updated_at or to_iso8601_utc(datetime.now(UTC)),
+        "generated_at": to_iso8601_utc(datetime.now(UTC)),
+    }
+
+
+async def get_public_tv_payload(db: AsyncSession) -> dict[str, Any]:
+    settings_payload = await get_tv_settings_payload(db)
+    collected_sources = await get_collected_tv_sources(db)
+    return build_public_tv_payload(settings_payload, collected_sources)
+
+
+def _extract_allowed_tv_hosts(sources: list[dict[str, Any]]) -> set[str]:
+    hosts: set[str] = set()
+    for source in sources:
+        for field in ("stream_url", "embed_url", "homepage_url", "youtube_channel"):
+            value = _clean_url(source.get(field))
+            if not value:
+                continue
+            parsed = urlparse(value)
+            if parsed.hostname:
+                hosts.add(parsed.hostname.lower())
+    return hosts
+
+
+def is_allowed_tv_proxy_url(url: str, sources: list[dict[str, Any]]) -> bool:
+    cleaned = _clean_url(url)
+    if not cleaned:
+        return False
+
+    parsed = urlparse(cleaned)
+    hostname = (parsed.hostname or "").lower()
+    if not hostname:
+        return False
+
+    allowed_hosts = _extract_allowed_tv_hosts(sources)
+    if hostname in allowed_hosts:
+        return True
+    return any(hostname.endswith(f".{allowed_host}") for allowed_host in allowed_hosts)

backend/tests/test_api.py

@@ -10,7 +10,12 @@ from app.core.config import settings
 from app.core.security import create_access_token
 from app.db.session import get_db
 from app.models.user import User
-from app.schemas.ai import AIProviderStatusResponse, SituationalAnalysisResponse
+from app.schemas.ai import (
+    AIProviderStatusResponse,
+    PlaygroundSessionResponse,
+    PlaygroundSessionState,
+    SituationalAnalysisResponse,
+)
 
 
 @pytest.fixture
@@ -30,6 +35,7 @@ async def test_health_check():
     data = response.json()
     assert data["status"] == "healthy"
     assert "version" in data
+    assert response.headers["x-request-id"]
 
 
 @pytest.mark.asyncio
@@ -90,6 +96,15 @@ async def test_alerts_without_auth():
     assert response.status_code == 401
 
 
+@pytest.mark.asyncio
+async def test_datasource_task_status_without_auth():
+    """Test datasource task-status endpoint requires authentication"""
+    transport = ASGITransport(app=app)
+    async with AsyncClient(transport=transport, base_url="http://test") as client:
+        response = await client.get("/api/v1/datasources/1/task-status")
+    assert response.status_code == 401
+
+
 @pytest.mark.asyncio
 async def test_alerts_endpoint_with_auth(auth_headers):
     """Test alerts endpoint with authentication"""
@@ -147,6 +162,345 @@ async def test_alerts_endpoint_with_auth(auth_headers):
         app.dependency_overrides.clear()
 
 
+@pytest.mark.asyncio
+async def test_system_log_sources_requires_super_admin(auth_headers):
+    def override_get_current_user():
+        return User(
+            id=1,
+            username="testuser",
+            email="test@example.com",
+            password_hash="hashed",
+            role="admin",
+            is_active=True,
+        )
+
+    app.dependency_overrides = {
+        __import__("app.core.security", fromlist=["get_current_user"]).get_current_user: override_get_current_user,
+    }
+    transport = ASGITransport(app=app)
+    try:
+        async with AsyncClient(transport=transport, base_url="http://test") as client:
+            response = await client.get("/api/v1/system/logs/sources", headers=auth_headers)
+        assert response.status_code == 403
+    finally:
+        app.dependency_overrides.clear()
+
+
+@pytest.mark.asyncio
+async def test_system_log_sources_with_super_admin(auth_headers):
+    def override_get_current_user():
+        return User(
+            id=1,
+            username="root",
+            email="root@example.com",
+            password_hash="hashed",
+            role="super_admin",
+            is_active=True,
+        )
+
+    app.dependency_overrides = {
+        __import__("app.core.security", fromlist=["get_current_user"]).get_current_user: override_get_current_user,
+    }
+    transport = ASGITransport(app=app)
+    try:
+        with patch(
+            "app.api.v1.system_control.list_log_sources",
+            return_value=[
+                {
+                    "source_id": "backend",
+                    "name": "后端服务",
+                    "kind": "file",
+                    "location": "/tmp/planet_backend.log",
+                    "description": "FastAPI 后端、调度器和采集任务共享日志。",
+                    "category": "service",
+                    "status": "ok",
+                }
+            ],
+        ):
+            async with AsyncClient(transport=transport, base_url="http://test") as client:
+                response = await client.get("/api/v1/system/logs/sources", headers=auth_headers)
+        assert response.status_code == 200
+        data = response.json()
+        assert data["items"][0]["source_id"] == "backend"
+    finally:
+        app.dependency_overrides.clear()
+
+
+@pytest.mark.asyncio
+async def test_system_log_snapshot_with_super_admin(auth_headers):
+    def override_get_current_user():
+        return User(
+            id=1,
+            username="root",
+            email="root@example.com",
+            password_hash="hashed",
+            role="super_admin",
+            is_active=True,
+        )
+
+    app.dependency_overrides = {
+        __import__("app.core.security", fromlist=["get_current_user"]).get_current_user: override_get_current_user,
+    }
+    transport = ASGITransport(app=app)
+    try:
+        with patch(
+            "app.api.v1.system_control.read_log_snapshot",
+            return_value={
+                "source_id": "backend",
+                "name": "后端服务",
+                "kind": "file",
+                "location": "/tmp/planet_backend.log",
+                "description": "FastAPI 后端、调度器和采集任务共享日志。",
+                "category": "service",
+                "status": "ok",
+                "level": "all",
+                "selected_levels": [],
+                "search_query": "",
+                "available_levels": ["all", "error", "warning", "info", "debug"],
+                "daily_markers": [],
+                "line_limit": 50,
+                "line_count": 2,
+                "lines": ["line 1", "line 2"],
+            },
+        ):
+            async with AsyncClient(transport=transport, base_url="http://test") as client:
+                response = await client.get("/api/v1/system/logs/backend?limit=50", headers=auth_headers)
+        assert response.status_code == 200
+        data = response.json()
+        assert data["source_id"] == "backend"
+        assert data["line_count"] == 2
+        assert data["lines"] == ["line 1", "line 2"]
+    finally:
+        app.dependency_overrides.clear()
+
+
+@pytest.mark.asyncio
+async def test_system_log_snapshot_supports_level_filter(auth_headers):
+    def override_get_current_user():
+        return User(
+            id=1,
+            username="root",
+            email="root@example.com",
+            password_hash="hashed",
+            role="super_admin",
+            is_active=True,
+        )
+
+    app.dependency_overrides = {
+        __import__("app.core.security", fromlist=["get_current_user"]).get_current_user: override_get_current_user,
+    }
+    transport = ASGITransport(app=app)
+    try:
+        with patch(
+            "app.api.v1.system_control.read_log_snapshot",
+            return_value={
+                "source_id": "backend",
+                "name": "后端服务",
+                "kind": "file",
+                "location": "/tmp/planet_backend.log",
+                "description": "FastAPI 后端、调度器和采集任务共享日志。",
+                "category": "service",
+                "status": "ok",
+                "level": "error",
+                "selected_levels": ["error"],
+                "search_query": "",
+                "available_levels": ["all", "error", "warning", "info", "debug"],
+                "daily_markers": [],
+                "line_limit": 50,
+                "line_count": 1,
+                "lines": ["ERROR: failed"],
+            },
+        ):
+            async with AsyncClient(transport=transport, base_url="http://test") as client:
+                response = await client.get("/api/v1/system/logs/backend?limit=50&level=error", headers=auth_headers)
+        assert response.status_code == 200
+        data = response.json()
+        assert data["level"] == "error"
+        assert data["lines"] == ["ERROR: failed"]
+    finally:
+        app.dependency_overrides.clear()
+
+
+@pytest.mark.asyncio
+async def test_system_log_snapshot_supports_date_range_filter(auth_headers):
+    def override_get_current_user():
+        return User(
+            id=1,
+            username="root",
+            email="root@example.com",
+            password_hash="hashed",
+            role="super_admin",
+            is_active=True,
+        )
+
+    app.dependency_overrides = {
+        __import__("app.core.security", fromlist=["get_current_user"]).get_current_user: override_get_current_user,
+    }
+    transport = ASGITransport(app=app)
+    try:
+        with patch(
+            "app.api.v1.system_control.read_log_snapshot",
+            return_value={
+                "source_id": "backend",
+                "name": "后端服务",
+                "kind": "file",
+                "location": "/tmp/planet_backend.log",
+                "description": "FastAPI 后端、调度器和采集任务共享日志。",
+                "category": "service",
+                "status": "ok",
+                "level": "all",
+                "selected_levels": [],
+                "search_query": "",
+                "available_levels": ["all", "error", "warning", "info", "debug"],
+                "daily_markers": [],
+                "line_limit": 50,
+                "line_count": 1,
+                "lines": ["2026-04-23 INFO: service started"],
+            },
+        ) as mock_read_log_snapshot:
+            async with AsyncClient(transport=transport, base_url="http://test") as client:
+                response = await client.get(
+                    "/api/v1/system/logs/backend?limit=50&start_date=2026-04-20&end_date=2026-04-23",
+                    headers=auth_headers,
+                )
+        assert response.status_code == 200
+        mock_read_log_snapshot.assert_called_once_with(
+            "backend",
+            50,
+            level="all",
+            levels=None,
+            start_date="2026-04-20",
+            end_date="2026-04-23",
+            search=None,
+        )
+    finally:
+        app.dependency_overrides.clear()
+
+
+@pytest.mark.asyncio
+async def test_system_log_snapshot_supports_levels_and_search_filter(auth_headers):
+    def override_get_current_user():
+        return User(
+            id=1,
+            username="root",
+            email="root@example.com",
+            password_hash="hashed",
+            role="super_admin",
+            is_active=True,
+        )
+
+    app.dependency_overrides = {
+        __import__("app.core.security", fromlist=["get_current_user"]).get_current_user: override_get_current_user,
+    }
+    transport = ASGITransport(app=app)
+    try:
+        with patch(
+            "app.api.v1.system_control.read_log_snapshot",
+            return_value={
+                "source_id": "backend",
+                "name": "后端服务",
+                "kind": "file",
+                "location": "/tmp/planet_backend.log",
+                "description": "FastAPI 后端、调度器和采集任务共享日志。",
+                "category": "service",
+                "status": "ok",
+                "level": "all",
+                "selected_levels": ["error", "warning"],
+                "search_query": "timeout",
+                "available_levels": ["all", "error", "warning", "info", "debug"],
+                "daily_markers": [],
+                "line_limit": 50,
+                "line_count": 1,
+                "lines": ["2026-04-23 10:00:00 ERROR timeout"],
+            },
+        ) as mock_read_log_snapshot:
+            async with AsyncClient(transport=transport, base_url="http://test") as client:
+                response = await client.get(
+                    "/api/v1/system/logs/backend?limit=50&levels=error,warning&search=timeout",
+                    headers=auth_headers,
+                )
+        assert response.status_code == 200
+        mock_read_log_snapshot.assert_called_once_with(
+            "backend",
+            50,
+            level="all",
+            levels="error,warning",
+            start_date=None,
+            end_date=None,
+            search="timeout",
+        )
+    finally:
+        app.dependency_overrides.clear()
+
+
+@pytest.mark.asyncio
+async def test_system_log_snapshot_rejects_invalid_date_range(auth_headers):
+    def override_get_current_user():
+        return User(
+            id=1,
+            username="root",
+            email="root@example.com",
+            password_hash="hashed",
+            role="super_admin",
+            is_active=True,
+        )
+
+    app.dependency_overrides = {
+        __import__("app.core.security", fromlist=["get_current_user"]).get_current_user: override_get_current_user,
+    }
+    transport = ASGITransport(app=app)
+    try:
+        async with AsyncClient(transport=transport, base_url="http://test") as client:
+            response = await client.get(
+                "/api/v1/system/logs/backend?start_date=2026-04-31",
+                headers=auth_headers,
+            )
+        assert response.status_code == 400
+        assert "start_date must be in YYYY-MM-DD format" in response.json()["detail"]
+    finally:
+        app.dependency_overrides.clear()
+
+
+@pytest.mark.asyncio
+async def test_ingest_earth_client_log_accepts_public_events():
+    transport = ASGITransport(app=app)
+    try:
+        with patch("app.api.v1.system_control.append_buffer_log") as mock_append_buffer_log:
+            with patch("app.api.v1.system_control.record_system_log", new_callable=AsyncMock) as mock_record_system_log:
+                async with AsyncClient(transport=transport, base_url="http://test") as client:
+                    response = await client.post(
+                        "/api/v1/system/logs/earth-client",
+                        json={
+                            "level": "error",
+                            "message": "登陆点加载失败: 登陆点接口返回 HTTP 500",
+                            "category": "startup-load",
+                            "module": "layer-startup",
+                        },
+                    )
+        assert response.status_code == 200
+        data = response.json()
+        assert data["accepted"] is True
+        assert data["source_id"] == "earth-client"
+        mock_append_buffer_log.assert_called_once()
+        mock_record_system_log.assert_awaited_once()
+        persisted_kwargs = mock_record_system_log.await_args.kwargs
+        assert persisted_kwargs["source"] == "earth-client"
+        assert persisted_kwargs["event"] == "earth.client.runtime_log"
+        assert persisted_kwargs["category"] == "startup-load"
+        assert persisted_kwargs["level"] == "error"
+    finally:
+        app.dependency_overrides.clear()
+
+
+@pytest.mark.asyncio
+async def test_request_id_header_is_echoed_when_provided():
+    transport = ASGITransport(app=app)
+    async with AsyncClient(transport=transport, base_url="http://test") as client:
+        response = await client.get("/health", headers={"X-Request-ID": "planet-test-request"})
+    assert response.status_code == 200
+    assert response.headers["x-request-id"] == "planet-test-request"
+
+
 @pytest.mark.asyncio
 async def test_invalid_token():
     """Test that invalid token is rejected"""
@@ -165,7 +519,8 @@ async def test_ai_provider_status_with_auth(auth_headers):
     class _FakeAIProviderClient:
         async def get_status(self, request_id=None):
             return AIProviderStatusResponse(
-                provider="openai_compatible",
+                provider="minimax",
+                api="anthropic-messages",
                 enabled=True,
                 configured=True,
                 model="test-model",
@@ -193,6 +548,7 @@ async def test_ai_provider_status_with_auth(auth_headers):
         assert response.status_code == 200
         data = response.json()
         assert "provider" in data
+        assert "api" in data
         assert "configured" in data
     finally:
         app.dependency_overrides.clear()
@@ -207,6 +563,9 @@ async def test_ai_situational_analysis_returns_503_when_disabled(auth_headers):
                 provider="openai_compatible",
                 model="test-model",
                 content="1) 态势摘要: 测试返回",
+                content_blocks=[],
+                text_blocks=["1) 态势摘要: 测试返回"],
+                thinking_blocks=[],
                 raw_response={"id": "mock-response"},
             )
 
@@ -241,5 +600,297 @@ async def test_ai_situational_analysis_returns_503_when_disabled(auth_headers):
         data = response.json()
         assert data["provider"] == "openai_compatible"
         assert data["content"]
+        assert "content_blocks" in data
+        assert "text_blocks" in data
+        assert "thinking_blocks" in data
+    finally:
+        app.dependency_overrides.clear()
+
+
+@pytest.mark.asyncio
+async def test_get_playground_session_with_auth(auth_headers):
+    """Test playground session restore endpoint."""
+
+    def override_get_current_user():
+        return User(
+            id=1,
+            username="testuser",
+            email="test@example.com",
+            password_hash="hashed",
+            role="admin",
+            is_active=True,
+        )
+
+    async def override_get_db():
+        yield AsyncMock()
+
+    app.dependency_overrides = {
+        __import__("app.core.security", fromlist=["get_current_user"]).get_current_user: override_get_current_user,
+        get_db: override_get_db,
+    }
+    transport = ASGITransport(app=app)
+    try:
+        with patch(
+            "app.api.v1.ai.get_playground_session",
+            new=AsyncMock(
+                return_value=PlaygroundSessionResponse(
+                    id="1",
+                    session_key="default",
+                    title="Playground 会话",
+                    state=PlaygroundSessionState(
+                        messages=[{"id": "msg-1", "role": "user", "content": "hello"}],
+                        title="测试标题",
+                        objective="测试目标",
+                    ),
+                    created_at="2026-04-10T00:00:00+00:00",
+                    updated_at="2026-04-10T00:00:00+00:00",
+                )
+            ),
+        ):
+            async with AsyncClient(transport=transport, base_url="http://test") as client:
+                response = await client.get("/api/v1/ai/playground/session", headers=auth_headers)
+        assert response.status_code == 200
+        data = response.json()
+        assert data["session_key"] == "default"
+        assert data["state"]["messages"][0]["content"] == "hello"
+    finally:
+        app.dependency_overrides.clear()
+
+
+@pytest.mark.asyncio
+async def test_save_playground_session_with_auth(auth_headers):
+    """Test playground session save endpoint."""
+
+    def override_get_current_user():
+        return User(
+            id=1,
+            username="testuser",
+            email="test@example.com",
+            password_hash="hashed",
+            role="admin",
+            is_active=True,
+        )
+
+    async def override_get_db():
+        yield AsyncMock()
+
+    app.dependency_overrides = {
+        __import__("app.core.security", fromlist=["get_current_user"]).get_current_user: override_get_current_user,
+        get_db: override_get_db,
+    }
+    transport = ASGITransport(app=app)
+    try:
+        with patch(
+            "app.api.v1.ai.upsert_playground_session",
+            new=AsyncMock(
+                return_value=PlaygroundSessionResponse(
+                    id="1",
+                    session_key="default",
+                    title="测试标题",
+                    state=PlaygroundSessionState(
+                        messages=[{"id": "msg-1", "role": "user", "content": "hello"}],
+                        title="测试标题",
+                        objective="测试目标",
+                    ),
+                    created_at="2026-04-10T00:00:00+00:00",
+                    updated_at="2026-04-10T00:00:00+00:00",
+                )
+            ),
+        ):
+            async with AsyncClient(transport=transport, base_url="http://test") as client:
+                response = await client.put(
+                    "/api/v1/ai/playground/session",
+                    headers=auth_headers,
+                    json={
+                        "session_key": "default",
+                        "title": "测试标题",
+                        "state": {
+                            "messages": [{"id": "msg-1", "role": "user", "content": "hello"}],
+                            "selectedPresetKey": "bgp-brief",
+                            "title": "测试标题",
+                            "objective": "测试目标",
+                            "constraints": "",
+                            "inputValue": "",
+                            "analysis": None,
+                            "latestAnalysisMessageId": None,
+                            "analysisMeta": {},
+                            "helpExpanded": True,
+                        },
+                    },
+                )
+        assert response.status_code == 200
+        data = response.json()
+        assert data["title"] == "测试标题"
+        assert data["state"]["objective"] == "测试目标"
+    finally:
+        app.dependency_overrides.clear()
+
+
+@pytest.mark.asyncio
+async def test_ai_bgp_brief_endpoint_persists_fact_snapshot(auth_headers):
+    class _FakeAIProviderClient:
+        async def analyze(self, _payload, request_id=None):
+            return SituationalAnalysisResponse(
+                provider="minimax",
+                model="MiniMax-M2.5",
+                content="# BGP AI 简报\n\n事实摘要：测试",
+                content_blocks=[],
+                text_blocks=["# BGP AI 简报\n\n事实摘要：测试"],
+                thinking_blocks=[],
+                raw_response={"id": "mock-bgp-brief"},
+            )
+
+    def override_get_current_user():
+        return User(
+            id=1,
+            username="testuser",
+            email="test@example.com",
+            password_hash="hashed",
+            role="admin",
+            is_active=True,
+        )
+
+    async def override_get_db():
+        yield AsyncMock()
+
+    async def _fake_build_bgp_brief_request(_db, **_kwargs):
+        request_payload = __import__("app.schemas.ai", fromlist=["SituationalAnalysisRequest"]).SituationalAnalysisRequest(
+            title="BGP 态势 AI 简报",
+            objective="生成值班简报",
+            observations=["事实A", "事实B"],
+            constraints=["不要编造"],
+            context={"incident_total": 2, "active_collectors": 3},
+        )
+        return request_payload, ["事实A", "事实B"], {"incident_total": 2, "active_collectors": 3}
+
+    app.dependency_overrides = {
+        __import__("app.core.security", fromlist=["get_current_user"]).get_current_user: override_get_current_user,
+        __import__("app.services.ai_client", fromlist=["get_ai_provider_client"]).get_ai_provider_client: lambda: _FakeAIProviderClient(),
+        get_db: override_get_db,
+    }
+    transport = ASGITransport(app=app)
+    try:
+        with patch("app.api.v1.ai.build_bgp_brief_request", side_effect=_fake_build_bgp_brief_request):
+            async with AsyncClient(transport=transport, base_url="http://test") as client:
+                response = await client.post("/api/v1/ai/bgp/brief", headers=auth_headers, json={})
+        assert response.status_code == 200
+        data = response.json()
+        assert data["facts"] == ["事实A", "事实B"]
+        assert data["context"]["incident_total"] == 2
+        assert data["content_markdown"]
+    finally:
+        app.dependency_overrides.clear()
+
+
+@pytest.mark.asyncio
+async def test_ai_alert_brief_endpoint_with_auth(auth_headers):
+    class _FakeAIProviderClient:
+        async def analyze(self, _payload, request_id=None):
+            return SituationalAnalysisResponse(
+                provider="minimax",
+                model="MiniMax-M2.7",
+                content="事实摘要：告警测试。风险研判：告警测试。建议动作：告警测试。",
+                content_blocks=[],
+                text_blocks=["事实摘要：告警测试。风险研判：告警测试。建议动作：告警测试。"],
+                thinking_blocks=[],
+                raw_response={"id": "mock-alert-brief"},
+            )
+
+    def override_get_current_user():
+        return User(
+            id=1,
+            username="testuser",
+            email="test@example.com",
+            password_hash="hashed",
+            role="admin",
+            is_active=True,
+        )
+
+    async def override_get_db():
+        yield AsyncMock()
+
+    async def _fake_build_alert_brief_request(_db, **_kwargs):
+        request_payload = __import__("app.schemas.ai", fromlist=["SituationalAnalysisRequest"]).SituationalAnalysisRequest(
+            title="告警态势 AI 简报",
+            objective="输出告警简报",
+            observations=["告警事实A", "告警事实B"],
+            constraints=["不要编造"],
+            context={"active_alerts": 3, "top_datasources": {"bgp": 2}},
+        )
+        return request_payload, ["告警事实A", "告警事实B"], {"active_alerts": 3, "top_datasources": {"bgp": 2}}
+
+    app.dependency_overrides = {
+        __import__("app.core.security", fromlist=["get_current_user"]).get_current_user: override_get_current_user,
+        __import__("app.services.ai_client", fromlist=["get_ai_provider_client"]).get_ai_provider_client: lambda: _FakeAIProviderClient(),
+        get_db: override_get_db,
+    }
+    transport = ASGITransport(app=app)
+    try:
+        with patch("app.api.v1.ai.build_alert_brief_request", side_effect=_fake_build_alert_brief_request):
+            async with AsyncClient(transport=transport, base_url="http://test") as client:
+                response = await client.post("/api/v1/ai/alerts/brief", headers=auth_headers, json={})
+        assert response.status_code == 200
+        data = response.json()
+        assert data["title"] == "告警态势 AI 简报"
+        assert data["facts"] == ["告警事实A", "告警事实B"]
+        assert data["context"]["active_alerts"] == 3
+        assert data["content"]
+    finally:
+        app.dependency_overrides.clear()
+
+
+@pytest.mark.asyncio
+async def test_ai_situational_alert_brief_endpoint_with_auth(auth_headers):
+    class _FakeAIProviderClient:
+        async def analyze(self, _payload, request_id=None):
+            return SituationalAnalysisResponse(
+                provider="minimax",
+                model="MiniMax-M2.7",
+                content="事实摘要：态势测试。风险研判：态势测试。建议动作：态势测试。",
+                content_blocks=[],
+                text_blocks=["事实摘要：态势测试。风险研判：态势测试。建议动作：态势测试。"],
+                thinking_blocks=[],
+                raw_response={"id": "mock-situational-brief"},
+            )
+
+    def override_get_current_user():
+        return User(
+            id=1,
+            username="testuser",
+            email="test@example.com",
+            password_hash="hashed",
+            role="admin",
+            is_active=True,
+        )
+
+    async def override_get_db():
+        yield AsyncMock()
+
+    async def _fake_build_situational_alert_brief_request(_db):
+        request_payload = __import__("app.schemas.ai", fromlist=["SituationalAnalysisRequest"]).SituationalAnalysisRequest(
+            title="态势告警 AI 简报",
+            objective="输出态势告警简报",
+            observations=["态势事实A", "态势事实B"],
+            constraints=["不要编造"],
+            context={"active_system_alerts": 2, "active_bgp_incidents": 1},
+        )
+        return request_payload, ["态势事实A", "态势事实B"], {"active_system_alerts": 2, "active_bgp_incidents": 1}
+
+    app.dependency_overrides = {
+        __import__("app.core.security", fromlist=["get_current_user"]).get_current_user: override_get_current_user,
+        __import__("app.services.ai_client", fromlist=["get_ai_provider_client"]).get_ai_provider_client: lambda: _FakeAIProviderClient(),
+        get_db: override_get_db,
+    }
+    transport = ASGITransport(app=app)
+    try:
+        with patch("app.api.v1.ai.build_situational_alert_brief_request", side_effect=_fake_build_situational_alert_brief_request):
+            async with AsyncClient(transport=transport, base_url="http://test") as client:
+                response = await client.post("/api/v1/ai/situational-alerts/brief", headers=auth_headers, json={})
+        assert response.status_code == 200
+        data = response.json()
+        assert data["title"] == "态势告警 AI 简报"
+        assert data["facts"] == ["态势事实A", "态势事实B"]
+        assert data["context"]["active_system_alerts"] == 2
+        assert data["content"]
     finally:
         app.dependency_overrides.clear()

backend/tests/test_earth_news.py

@@ -0,0 +1,49 @@
+from datetime import UTC, datetime
+
+from app.services.earth_news import ParsedNewsItem, _serialize_item
+
+
+def test_serialize_item_includes_region_anchor_for_cruise():
+    item = ParsedNewsItem(
+        id="google-apac:test",
+        title="Example APAC story",
+        summary="Example summary",
+        url="https://example.com/story",
+        source="Example Source",
+        feed_name="Global Monitor / APAC",
+        feed_region="asia-pacific",
+        homepage_url="https://example.com",
+        published_at=datetime(2026, 4, 23, 2, 30, tzinfo=UTC),
+    )
+
+    payload = _serialize_item(item, active_region="asia-pacific")
+
+    assert payload["latitude"] == 1.3521
+    assert payload["longitude"] == 103.8198
+    assert payload["location_label"] == "亚太"
+    assert payload["location_inferred"] is True
+    assert payload["is_focus_match"] is True
+    assert payload["published_at"] == "2026-04-23T02:30:00Z"
+
+
+def test_serialize_item_falls_back_to_global_anchor():
+    item = ParsedNewsItem(
+        id="custom:test",
+        title="Fallback story",
+        summary="Fallback summary",
+        url="https://example.com/fallback",
+        source="Fallback Source",
+        feed_name="Fallback Feed",
+        feed_region="unknown-region",
+        homepage_url="https://example.com",
+        published_at=None,
+    )
+
+    payload = _serialize_item(item, active_region="americas")
+
+    assert payload["latitude"] == 20.0
+    assert payload["longitude"] == 0.0
+    assert payload["location_label"] == "全球"
+    assert payload["location_inferred"] is True
+    assert payload["is_focus_match"] is False
+    assert payload["published_at"] is None

backend/tests/test_logging.py

@@ -0,0 +1,78 @@
+from __future__ import annotations
+
+import logging
+
+from io import StringIO
+
+from app.core.logging import PlanetContextFilter, PlanetFormatter, get_logger
+from app.core.request_context import set_request_id
+
+
+def _capture_output(callback):
+    stream = StringIO()
+    handler = logging.StreamHandler(stream)
+    handler.setFormatter(PlanetFormatter(datefmt="%Y-%m-%d %H:%M:%S"))
+    handler.addFilter(PlanetContextFilter())
+
+    adapter = get_logger("tests.logging")
+    target_logger = adapter.logger
+    original_handlers = list(target_logger.handlers)
+    original_level = target_logger.level
+    original_propagate = target_logger.propagate
+
+    target_logger.handlers = [handler]
+    target_logger.setLevel(logging.INFO)
+    target_logger.propagate = False
+
+    try:
+        callback(adapter)
+    finally:
+        handler.flush()
+        target_logger.handlers = original_handlers
+        target_logger.setLevel(original_level)
+        target_logger.propagate = original_propagate
+
+    return stream.getvalue()
+
+
+def test_structured_logger_injects_request_id_and_event():
+    set_request_id("req-test-123")
+    try:
+        output = _capture_output(
+            lambda logger: logger.info_event(
+                "collector started",
+                event="collector.run.started",
+                context={"collector_name": "bgp_news"},
+            )
+        )
+    finally:
+        set_request_id(None)
+
+    assert "request_id=req-test-123" in output
+    assert "event=collector.run.started" in output
+    assert "service=backend" in output
+    assert '"collector_name": "bgp_news"' in output
+
+
+def test_structured_logger_redacts_sensitive_text_and_context():
+    set_request_id("req-test-redact")
+    try:
+        output = _capture_output(
+            lambda logger: logger.error_event(
+                "Authorization: Bearer super-secret-token",
+                event="auth.token.failed",
+                context={
+                    "token": "plain-secret",
+                    "nested": {"password": "hunter2"},
+                    "safe": "visible",
+                },
+            )
+        )
+    finally:
+        set_request_id(None)
+
+    assert "super-secret-token" not in output
+    assert "plain-secret" not in output
+    assert "hunter2" not in output
+    assert "[REDACTED]" in output
+    assert '"safe": "visible"' in output

backend/tests/test_system_logs.py

@@ -0,0 +1,217 @@
+from __future__ import annotations
+
+import json
+
+from pathlib import Path
+
+from app.services import system_logs
+
+
+class FakeRedis:
+    def __init__(self) -> None:
+        self.store: dict[str, list[str]] = {}
+
+    def rpush(self, key: str, value: str) -> None:
+        self.store.setdefault(key, []).append(value)
+
+    def ltrim(self, key: str, start: int, end: int) -> None:
+        items = self.store.get(key, [])
+        normalized_end = None if end == -1 else end + 1
+        self.store[key] = items[start:normalized_end]
+
+    def expire(self, key: str, seconds: int) -> None:
+        return None
+
+    def lrange(self, key: str, start: int, end: int) -> list[str]:
+        items = self.store.get(key, [])
+        normalized_end = None if end == -1 else end + 1
+        return items[start:normalized_end]
+
+    def llen(self, key: str) -> int:
+        return len(self.store.get(key, []))
+
+
+def test_read_log_snapshot_uses_structured_buffer_timestamp_level_and_search(monkeypatch):
+    fake_redis = FakeRedis()
+    monkeypatch.setattr(system_logs, "redis_client", fake_redis)
+    monkeypatch.setattr(
+        system_logs,
+        "LOG_SOURCES",
+        {
+            "earth-client": system_logs.LogSource(
+                source_id="earth-client",
+                name="Earth 浏览器端",
+                kind="buffer",
+                location="redis://planet:system_logs:earth-client",
+                description="Earth 浏览器端上报日志",
+                category="client",
+                buffer_key=system_logs.get_buffer_log_key("earth-client"),
+            )
+        },
+    )
+
+    fake_redis.rpush(
+        system_logs.get_buffer_log_key("earth-client"),
+        json.dumps(
+            {
+                "timestamp": "2026-04-22T10:15:30Z",
+                "level": "warning",
+                "message": "news feed degraded",
+                "context": {"module": "news", "detail": "timeout"},
+            },
+            ensure_ascii=False,
+        ),
+    )
+    fake_redis.rpush(
+        system_logs.get_buffer_log_key("earth-client"),
+        json.dumps(
+            {
+                "timestamp": "2026-04-23T06:01:00Z",
+                "level": "error",
+                "message": "landing points failed",
+                "context": {"module": "layer-startup", "detail": "http 500"},
+            },
+            ensure_ascii=False,
+        ),
+    )
+
+    snapshot = system_logs.read_log_snapshot(
+        "earth-client",
+        50,
+        levels="error,warning",
+        start_date="2026-04-23",
+        end_date="2026-04-23",
+        search="landing",
+    )
+
+    assert snapshot is not None
+    assert snapshot["selected_levels"] == ["error", "warning"]
+    assert snapshot["search_query"] == "landing"
+    assert snapshot["line_count"] == 1
+    assert snapshot["lines"][0].startswith("2026-04-23 06:01:00 ERROR landing points failed")
+    assert snapshot["daily_markers"] == [
+        {"date_token": "2026-04-23", "total": 1, "dominant_level": "error"}
+    ]
+
+
+def test_read_log_snapshot_parses_file_timestamp_and_builds_markers(tmp_path: Path, monkeypatch):
+    log_path = tmp_path / "backend.log"
+    log_path.write_text(
+        "\n".join(
+            [
+                "2026-04-22 08:00:00 INFO service booted",
+                "2026-04-23 09:15:00 WARNING disk pressure detected",
+                "2026-04-23 09:16:00 ERROR sync failed",
+                "2026-04-24 10:00:00 DEBUG collector trace",
+            ]
+        ),
+        encoding="utf-8",
+    )
+
+    monkeypatch.setattr(
+        system_logs,
+        "LOG_SOURCES",
+        {
+            "backend": system_logs.LogSource(
+                source_id="backend",
+                name="后端服务",
+                kind="file",
+                location=str(log_path),
+                description="测试文件日志",
+                category="service",
+            )
+        },
+    )
+
+    snapshot = system_logs.read_log_snapshot(
+        "backend",
+        50,
+        levels="warning,error",
+        search="failed",
+    )
+
+    assert snapshot is not None
+    assert snapshot["line_count"] == 1
+    assert snapshot["lines"] == ["2026-04-23 09:16:00 ERROR sync failed"]
+    assert snapshot["daily_markers"] == [
+        {"date_token": "2026-04-23", "total": 1, "dominant_level": "error"}
+    ]
+    assert snapshot["status"] == "ok"
+
+
+def test_append_buffer_log_persists_normalized_level(monkeypatch):
+    fake_redis = FakeRedis()
+    monkeypatch.setattr(system_logs, "redis_client", fake_redis)
+
+    system_logs.append_buffer_log(
+        "earth-client",
+        level="warn",
+        message="feed delayed",
+        context={"module": "news"},
+    )
+
+    stored_items = fake_redis.lrange(system_logs.get_buffer_log_key("earth-client"), 0, -1)
+    payload = json.loads(stored_items[0])
+    assert payload["level"] == "warning"
+    assert payload["message"] == "feed delayed"
+
+
+def test_infer_log_level_prefers_leading_prefix_over_query_string():
+    line = 'INFO: 127.0.0.1 - "GET /api/v1/system/logs/backend?limit=200&level=error&levels=error HTTP/1.1" 200 OK'
+
+    entry = system_logs.parse_text_log_entry(line)
+
+    assert entry.level == "info"
+
+
+def test_parse_text_log_entry_does_not_promote_exception_context_to_error():
+    line = "websockets.exceptions.ConnectionClosedError: sent 1011 (internal error) keepalive ping timeout"
+
+    entry = system_logs.parse_text_log_entry(line)
+
+    assert entry.level is None
+
+
+def test_parse_text_log_entry_still_detects_explicit_error_prefix():
+    line = "ERROR: [Errno 98] Address already in use"
+
+    entry = system_logs.parse_text_log_entry(line)
+
+    assert entry.level == "error"
+
+
+def test_read_log_snapshot_strips_nul_bytes_from_file_lines(tmp_path: Path, monkeypatch):
+    log_path = tmp_path / "backend.log"
+    log_path.write_bytes(
+        (
+            b"INFO: service booted\n"
+            b"ERROR: bind failed\n"
+            + b"\x00" * 32
+            + b"2026-04-23 23:41:32 INFO service=backend message=request served\n"
+        )
+    )
+
+    monkeypatch.setattr(
+        system_logs,
+        "LOG_SOURCES",
+        {
+            "backend": system_logs.LogSource(
+                source_id="backend",
+                name="后端服务",
+                kind="file",
+                location=str(log_path),
+                description="测试文件日志",
+                category="service",
+            )
+        },
+    )
+
+    snapshot = system_logs.read_log_snapshot("backend", 50)
+
+    assert snapshot is not None
+    assert snapshot["line_count"] == 3
+    assert snapshot["lines"] == [
+        "INFO: service booted",
+        "ERROR: bind failed",
+        "2026-04-23 23:41:32 INFO service=backend message=request served",
+    ]

backend/tests/test_visualization_compute_centers.py

@@ -0,0 +1,347 @@
+from datetime import datetime, timezone
+
+import pytest
+from httpx import ASGITransport, AsyncClient
+
+from app.api.v1.visualization import convert_compute_centers_to_geojson
+from app.db.session import get_db
+from app.main import app
+from app.models.collected_data import CollectedData
+
+
+def _build_record(
+    *,
+    record_id: int,
+    source: str,
+    data_type: str,
+    name: str,
+    country: str,
+    city: str,
+    latitude: float,
+    longitude: float,
+    metadata: dict,
+):
+    return CollectedData(
+        id=record_id,
+        source=source,
+        data_type=data_type,
+        source_id=f"{source}-{record_id}",
+        name=name,
+        extra_data={
+            "country": country,
+            "city": city,
+            "latitude": latitude,
+            "longitude": longitude,
+            **metadata,
+        },
+        collected_at=datetime(2026, 4, 22, tzinfo=timezone.utc),
+        reference_date=datetime(2026, 4, 21, tzinfo=timezone.utc),
+        is_current=True,
+    )
+
+
+def test_convert_compute_centers_to_geojson_unifies_sources():
+    top500_record = _build_record(
+        record_id=1,
+        source="top500",
+        data_type="supercomputer",
+        name="Frontier",
+        country="United States",
+        city="Oak Ridge",
+        latitude=35.93,
+        longitude=-84.31,
+        metadata={
+            "rank": 1,
+            "manufacturer": "HPE",
+            "organization": "ORNL",
+            "rmax": 1102000.0,
+            "cores": 8730112,
+            "power": 21510.0,
+        },
+    )
+    gpu_record = _build_record(
+        record_id=2,
+        source="epoch_ai_gpu",
+        data_type="gpu_cluster",
+        name="Colossus",
+        country="United States",
+        city="Memphis",
+        latitude=35.15,
+        longitude=-90.05,
+        metadata={
+            "organization": "xAI",
+            "gpu_type": "H100",
+            "gpu_count": 100000,
+            "value": "20000",
+            "unit": "TFlop/s",
+        },
+    )
+
+    payload = convert_compute_centers_to_geojson([top500_record, gpu_record])
+
+    assert payload["type"] == "FeatureCollection"
+    assert len(payload["features"]) == 2
+
+    supercomputer_feature = payload["features"][0]
+    assert supercomputer_feature["properties"]["site_type"] == "supercomputer"
+    assert supercomputer_feature["properties"]["capacity_unit"] == "GFlops"
+    assert supercomputer_feature["properties"]["capacity_band"] == "exascale"
+    assert supercomputer_feature["properties"]["operator"] == "ORNL"
+    assert supercomputer_feature["properties"]["location_precision"] == "precise"
+    assert supercomputer_feature["properties"]["is_estimated"] is False
+
+    gpu_feature = payload["features"][1]
+    assert gpu_feature["properties"]["site_type"] == "gpu_cluster"
+    assert gpu_feature["properties"]["vendor"] == "H100"
+    assert gpu_feature["properties"]["gpu_count"] == 100000
+    assert gpu_feature["properties"]["capacity_band"] == "large"
+    assert gpu_feature["properties"]["location_precision"] == "precise"
+
+
+def test_convert_compute_centers_to_geojson_uses_coordinate_hints():
+    hinted_record = _build_record(
+        record_id=3,
+        source="top500",
+        data_type="supercomputer",
+        name="Frontier",
+        country="United States",
+        city="",
+        latitude=0.0,
+        longitude=0.0,
+        metadata={
+            "organization": "Oak Ridge National Laboratory",
+            "rmax": 1102000.0,
+        },
+    )
+
+    payload = convert_compute_centers_to_geojson([hinted_record])
+
+    assert len(payload["features"]) == 1
+    coords = payload["features"][0]["geometry"]["coordinates"]
+    assert coords[0] == pytest.approx(-84.3107)
+    assert coords[1] == pytest.approx(35.9319)
+    assert payload["features"][0]["properties"]["is_estimated"] is True
+    assert payload["features"][0]["properties"]["location_precision"] == "estimated_site"
+
+
+def test_convert_compute_centers_to_geojson_falls_back_to_country_centroid():
+    centroid_record = _build_record(
+        record_id=4,
+        source="epoch_ai_gpu",
+        data_type="gpu_cluster",
+        name="Unknown Cluster",
+        country="United States",
+        city="",
+        latitude=0.0,
+        longitude=0.0,
+        metadata={
+            "organization": "Unknown Operator",
+            "value": "10000",
+            "unit": "TFlop/s",
+        },
+    )
+
+    payload = convert_compute_centers_to_geojson([centroid_record])
+
+    assert len(payload["features"]) == 1
+    props = payload["features"][0]["properties"]
+    coords = payload["features"][0]["geometry"]["coordinates"]
+    assert coords[0] == pytest.approx(-98.5795)
+    assert coords[1] == pytest.approx(39.8283)
+    assert props["is_estimated"] is True
+    assert props["location_precision"] == "estimated_country"
+    assert props["geography_mode"] == "country_centroid"
+
+
+@pytest.mark.asyncio
+async def test_compute_centers_geojson_endpoint_returns_stats():
+    records = [
+        _build_record(
+            record_id=1,
+            source="top500",
+            data_type="supercomputer",
+            name="Frontier",
+            country="United States",
+            city="Oak Ridge",
+            latitude=35.93,
+            longitude=-84.31,
+            metadata={"rank": 1, "rmax": 1102000.0},
+        ),
+        _build_record(
+            record_id=2,
+            source="epoch_ai_gpu",
+            data_type="gpu_cluster",
+            name="Colossus",
+            country="United States",
+            city="Memphis",
+            latitude=35.15,
+            longitude=-90.05,
+            metadata={"value": "20000", "unit": "TFlop/s"},
+        ),
+    ]
+
+    class _ScalarResult:
+        def __init__(self, rows):
+            self._rows = rows
+
+        def scalars(self):
+            class _Scalars:
+                def __init__(self, rows):
+                    self._rows = rows
+
+                def all(self):
+                    return self._rows
+
+            return _Scalars(self._rows)
+
+    class _FakeSession:
+        async def execute(self, _query):
+            return _ScalarResult(records)
+
+    async def override_get_db():
+        yield _FakeSession()
+
+    app.dependency_overrides[get_db] = override_get_db
+    transport = ASGITransport(app=app)
+    try:
+        async with AsyncClient(transport=transport, base_url="http://test") as client:
+            response = await client.get("/api/v1/visualization/geo/compute-centers")
+
+        assert response.status_code == 200
+        data = response.json()
+        assert data["count"] == 2
+        assert data["stats"]["supercomputers"] == 1
+        assert data["stats"]["gpu_clusters"] == 1
+        assert data["features"][0]["properties"]["data_type"] == "compute_center"
+    finally:
+        app.dependency_overrides.clear()
+
+
+@pytest.mark.asyncio
+async def test_visualization_geo_summary_returns_counts(monkeypatch):
+    records = [
+        _build_record(
+            record_id=1,
+            source="arcgis_cables",
+            data_type="submarine_cable",
+            name="Test Cable",
+            country="",
+            city="",
+            latitude=0,
+            longitude=0,
+            metadata={
+                "route_coordinates": [[[0, 0], [1, 1]]],
+                "status": "active",
+            },
+        ),
+        _build_record(
+            record_id=2,
+            source="arcgis_landing_points",
+            data_type="landing_point",
+            name="Test Landing",
+            country="United States",
+            city="New York",
+            latitude=40.7,
+            longitude=-74.0,
+            metadata={"city_id": 10},
+        ),
+        _build_record(
+            record_id=3,
+            source="celestrak_tle",
+            data_type="satellite_tle",
+            name="TESTSAT",
+            country="",
+            city="",
+            latitude=0,
+            longitude=0,
+            metadata={
+                "norad_cat_id": 12345,
+                "tle_line1": "1 12345U 98067A   24001.00000000  .00000000  00000-0  00000-0 0  9991",
+                "tle_line2": "2 12345  51.6000 100.0000 0001000  10.0000  20.0000 15.50000000    01",
+            },
+        ),
+        _build_record(
+            record_id=4,
+            source="top500",
+            data_type="supercomputer",
+            name="Frontier",
+            country="United States",
+            city="Oak Ridge",
+            latitude=35.93,
+            longitude=-84.31,
+            metadata={"rank": 1, "rmax": 1102000.0},
+        ),
+        _build_record(
+            record_id=5,
+            source="epoch_ai_gpu",
+            data_type="gpu_cluster",
+            name="Colossus",
+            country="United States",
+            city="Memphis",
+            latitude=35.15,
+            longitude=-90.05,
+            metadata={"value": "20000", "unit": "TFlop/s"},
+        ),
+    ]
+
+    class _ScalarResult:
+        def __init__(self, rows=None, scalar_value=None):
+            self._rows = rows or []
+            self._scalar_value = scalar_value
+
+        def scalar(self):
+            return self._scalar_value
+
+        def scalars(self):
+            class _Scalars:
+                def __init__(self, rows):
+                    self._rows = rows
+
+                def all(self):
+                    return self._rows
+
+            return _Scalars(self._rows)
+
+    class _FakeSession:
+        async def execute(self, query):
+            query_text = str(query)
+            if "bgp_incidents" in query_text:
+                return _ScalarResult(scalar_value=2)
+            if "bgp_anomalies" in query_text:
+                return _ScalarResult(scalar_value=3)
+            return _ScalarResult(rows=records)
+
+    async def override_get_db():
+        yield _FakeSession()
+
+    async def _fake_build_bgp_collector_coverage(*_args, **_kwargs):
+        return [
+            {"collector": "rrc00"},
+            {"collector": "rrc01"},
+        ]
+
+    monkeypatch.setattr(
+        "app.api.v1.visualization.build_bgp_collector_coverage",
+        _fake_build_bgp_collector_coverage,
+    )
+
+    app.dependency_overrides[get_db] = override_get_db
+    transport = ASGITransport(app=app)
+    try:
+        async with AsyncClient(transport=transport, base_url="http://test") as client:
+            response = await client.get("/api/v1/visualization/geo/summary")
+
+        assert response.status_code == 200
+        stats = response.json()["stats"]
+        assert stats["cable_count"] == 1
+        assert stats["landing_point_count"] == 1
+        assert stats["satellite_count"] == 1
+        assert stats["compute_center_count"] == 2
+        assert stats["supercomputer_count"] == 1
+        assert stats["gpu_cluster_count"] == 1
+        assert stats["bgp_event_count"] == 2
+        assert stats["bgp_incident_count"] == 2
+        assert stats["bgp_anomaly_count"] == 3
+        assert stats["bgp_collector_count"] == 2
+    finally:
+        app.dependency_overrides.clear()

docker-compose.yml

@@ -5,6 +5,8 @@ services:
     build:
       context: .
       dockerfile: aiprovider/Dockerfile
+    env_file:
+      - ./aiprovider/.env
     container_name: planet_aiprovider
     ports:
       - "8010:8010"

docs/deprecated/README.md

@@ -0,0 +1,22 @@
+# Deprecated Docs
+
+这个目录用于存放两类文档：
+
+1. 已经完成、主要保留为历史记录的实施计划
+2. 已被现有实现或新方案替代的旧计划
+
+放到这里并不代表这些文档“错误”，而是表示：
+
+- 它们不再适合作为当前开发的主指导文档
+- 如果要了解历史决策、演进路径或旧设计背景，仍然可以参考
+
+当前归档原则：
+
+- 明确写明“已完成”的计划，优先归档
+- 已被正式实现替代、继续放在 `docs/` 根目录会误导后续开发的计划，归档
+- 仍然指导未来开发、尚未完成或仍有明确执行价值的文档，继续保留在 `docs/`
+
+补充说明：
+
+- 一部分归档文档来自外部或临时工作流草案，例如 sisyphus 生成的初稿
+- 这类文档如果有可用内容，应先吸收到 `docs/plans/` 或 `docs/technical/`，再归档保留来源记录

docs/deprecated/earth-tv-live-module-plan.md

@@ -0,0 +1,117 @@
+# Earth 电视直播模块计划
+
+## 目标
+
+为 `Earth` 页面增加一个可配置、可扩展、可拖拽的电视直播模块：
+
+- 后台可配置新闻直播源
+- 默认兜底源为央视 `CCTV-4`
+- 未来可通过采集器接入世界各地新闻直播源
+- Earth 工具栏 `显示控制` 子菜单新增电视按钮
+- 点击后打开一个与其他 HUD 一致的可拖拽/可关闭窗口
+- 窗口内部可播放或承载新闻直播页面
+
+## 设计原则
+
+- 第一阶段先交付“后台可配 + Earth 可用 + 默认可回退”的版本
+- 公开读取接口与后台管理接口分离
+- 手工配置源与采集器源共用统一的前端消费结构
+- Earth 里的电视窗口必须复用现有 HUD 拖拽、关闭、布局最大化逻辑
+- 小屏下优先保证窗口完整显示，超出部分在窗口内部滚动
+
+## 分阶段实现
+
+### Phase 1：后端配置与公开读取
+
+- 在系统设置中新增 `tv` 分类
+- 定义直播源配置结构：
+  - `default_source_id`
+  - `auto_fallback`
+  - `sources[]`
+- 每个直播源至少包含：
+  - `id`
+  - `name`
+  - `provider`
+  - `region`
+  - `language`
+  - `source_type`
+  - `embed_url`
+  - `stream_url`
+  - `homepage_url`
+  - `is_enabled`
+  - `is_fallback`
+  - `sort_order`
+  - `collector_source`
+  - `notes`
+- 默认兜底源使用央视官网 `CCTV-4` 直播页
+- 新增公开读取接口，供 Earth 页面无登录态读取直播源配置
+
+### Phase 2：采集器扩展位
+
+- 新增 `news_live_streams` collector 占位
+- 规范采集器入库数据结构，使其能与后台手工配置源合并
+- TV 公开接口支持合并：
+  - 后台手工配置源
+  - 采集器入库源
+- 保持手工配置源优先级更高，避免采集器覆盖人工兜底配置
+
+### Phase 3：后台配置界面
+
+- 在系统配置页新增 `电视直播` tab
+- 支持：
+  - 查看当前默认源
+  - 开关自动回退
+  - 新增直播源
+  - 编辑直播源
+  - 删除直播源
+  - 启用/禁用直播源
+  - 将某个直播源设为默认源
+- 明确区分：
+  - 手工配置源
+  - 采集器来源
+
+### Phase 4：Earth HUD 集成
+
+- 在 `显示控制` 子菜单加入电视按钮
+- 新增 TV HUD 面板：
+  - 可拖拽
+  - 可关闭
+  - 支持显示/隐藏状态同步
+  - 参与布局最大化与恢复布局
+- 面板内容至少包含：
+  - 当前频道标题
+  - 源切换下拉菜单
+  - 刷新按钮
+  - 打开官网按钮
+  - 播放区域
+
+### Phase 5：播放策略
+
+- 第一版优先支持 `iframe`/嵌入页类直播源
+- 为未来扩展保留：
+  - `hls`
+  - `video`
+  - `external`
+- 如果默认源不可用：
+  - 优先回退到标记为 `is_fallback=true` 的源
+  - 若无明确回退源，则回退到第一个可用源
+- 面板内要有清晰的加载、错误、回退提示
+
+### Phase 6：打磨与清理
+
+- 统一 HUD 风格
+- 小屏下限制窗口尺寸并启用内部滚动
+- 避免窗口超出屏幕
+- 补最小验证
+- 清理临时代码、重复样式和无用资源
+
+## 首版交付定义
+
+当以下条件满足时，认为首版可用：
+
+- 后台可以配置新闻直播源
+- Earth 可以读取并显示默认直播源
+- 工具栏可打开电视窗口
+- 电视窗口可拖拽、可关闭
+- 央视 `CCTV-4` 作为默认兜底源可被使用
+- 代码结构已为后续采集器接入预留统一接口

docs/deprecated/hud-panel-component-plan.md

@@ -0,0 +1,165 @@
+# HUD Panel Component Plan
+
+## Goal
+
+Unify Earth HUD panels into a reusable component layer so new panels can share:
+
+- a consistent shell
+- a consistent header
+- a consistent action-button system
+- a consistent body and collapse pattern
+
+## Scope
+
+Target panels:
+
+- `tv-panel`
+- `news-panel`
+- `legend`
+- `layer-panel`
+- `earth-stats`
+- `info-card`
+- settings modal header/actions
+
+## Component Model
+
+### Base shell
+
+- `.hud-panel`
+- `.hud-panel--compact`
+- `.hud-panel--media`
+- `.hud-panel--collapsed`
+- `.hud-panel-hidden`
+- `.hud-panel.is-dragging`
+- `.hud-panel.is-layout-animating`
+
+### Header
+
+- `.hud-panel__header`
+- `.hud-panel__title-group`
+- `.hud-panel__title`
+- `.hud-panel__subtitle`
+- `.hud-panel__chip`
+- `.hud-panel__actions`
+
+Header baseline rule:
+
+- Header title styling is fixed by the component layer and should not drift per panel
+- Title font size, font weight, letter spacing, line height, text color, and vertical alignment come from the shared header tokens and structure
+- Header divider, border treatment, inner spacing, and title-to-actions alignment are part of the same shared baseline
+- Panel-specific header differences should be limited to explicit variants such as `compact` or `media`, or token overrides with documented intent
+- “Looks close enough” local header overrides should be treated as temporary compatibility code and removed during migration
+
+### Actions
+
+- `.hud-panel__action`
+- `.hud-panel__action--icon`
+- `.hud-panel__action--collapse`
+- `.hud-panel__action--close`
+- `.hud-panel__action--refresh`
+- `.hud-panel__action--external`
+
+Action-button baseline rule:
+
+- Header action buttons must have one fixed default style baseline across all HUD panels
+- Default width behavior, padding, icon size, radius, alignment, hover, and active feedback all come from `.hud-panel__action`
+- Panel-specific differences must be expressed through explicit variants or token overrides, not ad-hoc local button rewrites
+- `close` buttons are part of the same default action system and must not silently fall back to a separate legacy box model
+
+### Body
+
+- `.hud-panel__body`
+- `.hud-panel__body--scroll`
+- `.hud-panel__body--collapsible`
+
+### Collapse behavior
+
+- `.hud-panel--collapsed`
+- `.hud-panel--expand-up`
+- `.hud-panel--expand-down`
+
+Adaptive collapse / expand rule:
+
+- HUD panels support two expansion directions:
+  - top-to-bottom expansion
+  - bottom-to-top expansion
+- Expansion direction should be decided at runtime from available viewport space rather than hardcoded per panel
+- Use:
+  - `d` = available distance from the header anchor to the viewport bottom edge
+  - `h` = expected expanded panel height
+  - buffer = `20px`
+- Collapsed-state direction rule:
+  - if `d > h + 20px`, the next action direction is `expand-up`
+  - if `d <= h + 20px`, the next action direction is `expand-down`
+- To avoid jitter around the threshold, the shared controller should keep a small hysteresis band:
+  - if the current direction is already `up`, keep it until `d <= h`
+  - if the current direction is already `down`, keep it until `d > h + 20px`
+- The opposite edge is still a safety guard:
+  - if the chosen side cannot fit at all, fall back to the other side if it can fit
+  - if neither side fully fits, choose the side with more space and let the body scroll
+- If neither direction fully fits, choose the direction with more available space and let the body scroll
+- Collapse icon direction must match the active expansion direction so the icon always describes the real open/close motion
+- The collapse icon describes the next action, not the current state
+- This mapping is fixed component behavior and must not drift per panel:
+  - collapsed + expand-down => `expand_more`
+  - expanded + expand-down => `expand_less`
+  - collapsed + expand-up => `expand_less`
+  - expanded + expand-up => `expand_more`
+- Panels must not combine icon-name swapping with extra CSS rotation for the same collapse control
+- Expansion direction and icon direction must come from one shared source of truth in the component controller
+- The direction decision should be recomputed when opening, resizing the viewport, or restoring a dragged panel near another edge
+
+## Tokens
+
+Promote panel differences into CSS variables instead of duplicating selectors:
+
+- `--hud-panel-padding`
+- `--hud-header-padding`
+- `--hud-header-gap`
+- `--hud-action-padding`
+- `--hud-action-gap`
+- `--hud-action-icon-size`
+- `--hud-body-gap`
+- `--hud-body-max-height`
+- `--hud-chip-radius`
+- `--hud-title-font-size`
+- `--hud-title-font-weight`
+- `--hud-title-letter-spacing`
+- `--hud-title-line-height`
+- `--hud-title-color`
+- `--hud-header-border-color`
+- `--hud-header-divider-opacity`
+- `--hud-expand-direction`
+
+## Migration Order
+
+1. Build the shared component layer in `frontend/public/earth/css/hud.css`
+2. Migrate `tv-panel` and `news-panel` first as the reference implementation
+3. Migrate `legend` and `layer-panel` into a compact variant
+4. Migrate `earth-stats` and `info-card`
+5. Align settings modal header/actions with the same action system
+6. Remove legacy one-off button selectors after verification
+
+## Guardrails
+
+- Do not change panel behavior and data flow during the first pass
+- Keep old class names temporarily as compatibility hooks
+- Prefer variable overrides over per-panel reimplementation
+- Treat header action-button default styling as fixed component API, not per-panel design space
+- Treat header title typography, border, and divider styling as fixed component API, not per-panel design space
+- Treat collapse direction as a component behavior contract, not a one-off panel trick
+- Treat collapse icon semantics as a component behavior contract, not a per-panel visual preference
+- Verify header alignment and drag/collapse behavior after each migration batch
+
+## First Implementation Batch
+
+Batch 1 should only do:
+
+- shared header structure
+- shared action-button system
+- shared title typography and header border/divider baseline
+- shared collapsible body pattern
+- adaptive collapse direction logic and direction-aware collapse icons
+- migration of `tv-panel` and `news-panel`
+
+That keeps risk low while giving the rest of the HUD a stable target to migrate toward.

docs/deprecated/sisyphus-earth-architecture-refactor.md

@@ -1,3 +1,5 @@
+> Archived note: this document was originally created by sisyphus and later reviewed against the main docs set. Useful content has been absorbed into `docs/plans/` where appropriate.
+
 # 地球3D可视化架构重构计划
 
 ## 背景

docs/deprecated/sisyphus-predicted-orbit.md

@@ -1,3 +1,5 @@
+> Archived note: this document was originally created by sisyphus and later reviewed against the main docs set. Useful content has been absorbed into `docs/plans/` where appropriate.
+
 # 卫星预测轨道显示功能
 
 ## TL;DR

docs/deprecated/sisyphus-ue5-client.md

@@ -1,3 +1,5 @@
+> Archived note: this document was originally created by sisyphus and later reviewed against the main docs set. Useful content has been absorbed into `docs/plans/` where appropriate.
+
 # UE5 3D 大屏客户端开发计划
 
 ## 项目概述

docs/deprecated/sisyphus-webgl-instancing-satellites.md

@@ -1,3 +1,5 @@
+> Archived note: this document was originally created by sisyphus and later reviewed against the main docs set. Useful content has been absorbed into `docs/plans/` where appropriate.
+
 # WebGL Instancing 卫星渲染优化计划
 
 ## 背景

docs/plans/README.md

@@ -0,0 +1,38 @@
+# Plans Docs
+
+这里放“未来实施方案和未完成计划”的文档，重点回答：
+
+- 我们准备做什么
+- 为什么要做
+- 分几期做
+- 当前差距和下一步是什么
+
+适合放入这里的内容：
+
+- Earth / BGP / 地形 / 天球实施方案
+- AI Playground 发展计划
+- backend / datasource / agent roadmap
+- UE5 MVP 方案
+
+当前重点入口：
+
+- [earth-mobile-drawer-ui-plan.md](/home/ray/dev/linkong/planet/docs/plans/earth-mobile-drawer-ui-plan.md)
+- [earth-compute-center-bgp-style-plan.md](/home/ray/dev/linkong/planet/docs/plans/earth-compute-center-bgp-style-plan.md)
+- [earth-renderer-architecture-separation-plan.md](/home/ray/dev/linkong/planet/docs/plans/earth-renderer-architecture-separation-plan.md)
+- [earth-country-boundary-overlay-plan.md](/home/ray/dev/linkong/planet/docs/plans/earth-country-boundary-overlay-plan.md)
+- [earth-predicted-orbit-plan.md](/home/ray/dev/linkong/planet/docs/plans/earth-predicted-orbit-plan.md)
+- [earth-webgl-instancing-satellites-plan.md](/home/ray/dev/linkong/planet/docs/plans/earth-webgl-instancing-satellites-plan.md)
+- [earth-real-terrain-plan.md](/home/ray/dev/linkong/planet/docs/plans/earth-real-terrain-plan.md)
+- [earth-news-source-configuration-and-collector-plan.md](/home/ray/dev/linkong/planet/docs/plans/earth-news-source-configuration-and-collector-plan.md)
+- [frontend-ai-playground-development-plan.md](/home/ray/dev/linkong/planet/docs/plans/frontend-ai-playground-development-plan.md)
+- [ue5-mvp-fused-plan.md](/home/ray/dev/linkong/planet/docs/plans/ue5-mvp-fused-plan.md)
+
+不适合放入这里的内容：
+
+- 当前代码结构说明
+- 组件现状和实现入口
+- 已经落地的技术上下文说明
+
+这些应放入：
+
+- [docs/technical/README.md](/home/ray/dev/linkong/planet/docs/technical/README.md)

docs/plans/agents-agent-architecture-plan.md

@@ -0,0 +1,647 @@
+# Agent Architecture Plan
+
+## Overview
+
+This document defines the agent architecture for Planet.
+
+The architecture is intentionally broader than datasource health checking.
+
+It is designed to support both:
+
+- datasource health governance
+- future situational-awareness workflows
+
+The core idea is to avoid building a one-off "repair broken API links" agent.
+
+Instead, Planet should grow a reusable agent runtime that can:
+
+- collect evidence
+- evaluate signals
+- reason over incomplete information
+- generate proposals
+- produce assessments
+- execute limited actions under policy
+
+
+## Design Goal
+
+Build an agent foundation that can evolve in this order:
+
+1. datasource health checks
+2. datasource repair proposals
+3. signal correlation
+4. situational assessments
+5. controlled runtime actions
+
+This means the architecture should treat datasource health as one use case of the larger agent system, not as the whole system.
+
+
+## Core Principles
+
+1. Separate evidence from reasoning
+
+- raw signals should be gathered first
+- deterministic checks should run before LLM reasoning
+
+2. Agents do not own the defaults
+
+- repository defaults remain human-owned
+- agents operate on runtime state, proposals, and overrides
+
+3. Reasoning and action are different responsibilities
+
+- many agents should be read-only or propose-only
+- only tightly controlled flows may apply changes
+
+4. Shared runtime, specialized roles
+
+- multiple agent roles should share the same object model and orchestration patterns
+- health and situational-awareness agents should not invent incompatible payloads
+
+5. Auditability is mandatory
+
+- every proposal, assessment, and applied action should be attributable
+
+
+## System Layers
+
+Planet agent architecture should be split into four layers.
+
+### 1. Signal Layer
+
+Purpose:
+
+- gather raw evidence from internal and external systems
+
+Example sources:
+
+- collector outputs
+- datasource health checks
+- logs
+- snapshots
+- alerts
+- web search results
+- scraped pages
+- external APIs
+- operator inputs
+
+Responsibilities:
+
+- fetch
+- normalize
+- timestamp
+- tag with source and trust level
+
+This layer should not make high-level judgments.
+
+
+### 2. Evaluation Layer
+
+Purpose:
+
+- perform deterministic analysis
+
+Examples:
+
+- reachability checks
+- schema validation
+- threshold checks
+- time-window comparisons
+- anomaly counters
+- completeness checks
+
+Responsibilities:
+
+- classify signals into machine-readable findings
+- attach deterministic evidence
+
+This layer should avoid LLM dependency whenever possible.
+
+
+### 3. Reasoning Layer
+
+Purpose:
+
+- use LLMs when semantic interpretation or incomplete-information reasoning is needed
+
+Examples:
+
+- endpoint migration inference
+- multi-source event correlation
+- causality hypotheses
+- ambiguity reduction
+- assessment narrative generation
+- action recommendation generation
+
+Responsibilities:
+
+- synthesize evidence
+- produce hypotheses
+- rank confidence
+- explain reasoning boundaries
+
+This is the main place where `aiprovider` and web search are used.
+
+
+### 4. Action Layer
+
+Purpose:
+
+- convert proposals or assessments into controlled system actions
+
+Examples:
+
+- create runtime override
+- create proposal
+- publish alert
+- update operator task queue
+- generate summary artifact
+- trigger follow-up verification
+
+Responsibilities:
+
+- enforce policy
+- enforce approval requirements
+- verify post-action outcomes
+- record audit trails
+
+
+## Architecture Sketch
+
+```mermaid
+flowchart TD
+    A["Collectors / Logs / Snapshots / External APIs"] --> B["Signal Layer"]
+    W["Web Search / Page Fetch / Docs"] --> B
+    B --> C["Evaluation Layer"]
+    C --> D["Findings"]
+    D --> E["Reasoning Layer (LLM + Tools)"]
+    E --> F["Proposals"]
+    E --> G["Assessments"]
+    F --> H["Action Layer"]
+    H --> I["Runtime Overrides / Alerts / Tasks"]
+    H --> J["Verification Loop"]
+    J --> B
+
+    K["Policy Engine"] --> H
+    L["Audit / History Store"] --> H
+    L --> E
+    L --> C
+```
+
+
+## Agent Roles
+
+The first version should define these logical roles.
+
+### 1. Health Agent
+
+Primary use case:
+
+- datasource health governance
+
+Inputs:
+
+- datasource metadata
+- current endpoint
+- latest health records
+- latest failures
+- deterministic findings
+
+Outputs:
+
+- health interpretation
+- repair proposal
+- confidence
+- evidence references
+
+Typical action level:
+
+- propose-only
+
+
+### 2. Correlation Agent
+
+Primary use case:
+
+- identify whether multiple signals describe the same event or related events
+
+Inputs:
+
+- findings from multiple collectors
+- time windows
+- region / ASN / prefix / cable relationships
+- prior incidents
+
+Outputs:
+
+- grouped event candidates
+- correlation rationale
+- confidence per relationship
+
+Typical action level:
+
+- read-only
+
+
+### 3. Assessment Agent
+
+Primary use case:
+
+- produce situational-awareness outputs
+
+Inputs:
+
+- grouped events
+- findings
+- current context
+- historical context
+- operator constraints
+
+Outputs:
+
+- structured assessment
+- risk summary
+- evidence-backed recommendations
+- missing-information list
+
+Typical action level:
+
+- read-only or propose-only
+
+
+### 4. Recovery Agent
+
+Primary use case:
+
+- carry low-risk proposals into controlled runtime actions
+
+Inputs:
+
+- approved proposal
+- policy constraints
+- trusted-domain rules
+- verification checks
+
+Outputs:
+
+- applied override
+- failed application
+- rollback request
+
+Typical action level:
+
+- apply-limited
+
+
+## Shared Object Model
+
+All agents should work on a shared object model.
+
+That prevents the health subsystem and situational-awareness subsystem from drifting into incompatible payloads.
+
+### Signal
+
+Represents a raw observed fact.
+
+Examples:
+
+- a datasource returned HTTP 404
+- a collector returned empty results
+- BGP updates spiked in one region
+- a known endpoint now redirects elsewhere
+
+Suggested shape:
+
+```json
+{
+  "id": "sig_123",
+  "type": "datasource.http_failure",
+  "source": "ris_live_bgp",
+  "occurred_at": "2026-04-08T10:00:00Z",
+  "severity": "medium",
+  "payload": {},
+  "trust": 0.95
+}
+```
+
+
+### Finding
+
+Represents a deterministic or semi-deterministic interpretation of one or more signals.
+
+Examples:
+
+- `schema_changed`
+- `endpoint_unreachable`
+- `data_volume_abnormally_low`
+- `event_cluster_detected`
+
+Suggested shape:
+
+```json
+{
+  "id": "find_123",
+  "type": "datasource.schema_changed",
+  "source_ids": ["sig_123"],
+  "confidence": 0.92,
+  "evidence": [],
+  "details": {}
+}
+```
+
+
+### Proposal
+
+Represents a recommended action, not an already-applied action.
+
+Examples:
+
+- switch endpoint to new URL
+- disable bad override
+- escalate issue for manual review
+
+Suggested shape:
+
+```json
+{
+  "id": "prop_123",
+  "kind": "endpoint_override",
+  "target": "telegeography_cables",
+  "confidence": 0.84,
+  "reason": "Official docs now point to a new API path",
+  "payload": {},
+  "evidence_urls": [],
+  "status": "proposed"
+}
+```
+
+
+### Assessment
+
+Represents a structured situational-awareness output for operators or downstream systems.
+
+Examples:
+
+- current network posture summary
+- incident impact assessment
+- risk and response recommendations
+
+Suggested shape:
+
+```json
+{
+  "id": "assess_123",
+  "scope": "regional-network",
+  "risk_level": "high",
+  "summary": "Regional routing instability is increasing.",
+  "key_risks": [],
+  "evidence": [],
+  "recommendations": [],
+  "missing_data": []
+}
+```
+
+
+## State Machine
+
+The shared orchestration flow should look like this:
+
+```mermaid
+stateDiagram-v2
+    [*] --> Collect
+    Collect --> Validate
+    Validate --> Classify
+    Classify --> Reason
+    Reason --> Propose
+    Reason --> Assess
+    Propose --> Review
+    Review --> Apply
+    Apply --> Verify
+    Verify --> Archive
+    Assess --> Archive
+    Archive --> [*]
+```
+
+Definitions:
+
+- `Collect`: gather signals
+- `Validate`: run deterministic checks
+- `Classify`: create findings
+- `Reason`: invoke LLM reasoning when needed
+- `Propose`: create change proposals
+- `Review`: policy or human approval
+- `Apply`: perform limited runtime action
+- `Verify`: confirm action effect
+- `Archive`: store artifacts and decisions
+
+
+## Permission Model
+
+Each agent role should be assigned one of these action levels.
+
+### `read-only`
+
+Allowed:
+
+- read signals
+- search web
+- fetch pages
+- read internal state
+- generate findings and assessments
+
+Not allowed:
+
+- mutate config
+- write overrides
+- change live runtime behavior
+
+
+### `propose-only`
+
+Allowed:
+
+- everything in `read-only`
+- create proposals
+- create review tasks
+
+Not allowed:
+
+- apply live changes
+
+
+### `apply-limited`
+
+Allowed:
+
+- everything in `propose-only`
+- write approved runtime overrides
+- trigger verification checks
+
+Not allowed:
+
+- mutate repository defaults
+- make destructive data changes
+- bypass policy engine
+
+
+## Runtime Components
+
+The first durable architecture should introduce these components.
+
+### 1. Signal Store
+
+Stores normalized evidence and health outputs.
+
+
+### 2. Finding Store
+
+Stores deterministic classifications that can be reused by multiple agents.
+
+
+### 3. Proposal Store
+
+Stores recommended actions with evidence and confidence.
+
+
+### 4. Assessment Store
+
+Stores structured situational-awareness outputs.
+
+
+### 5. Policy Engine
+
+Decides:
+
+- whether agent may run
+- whether proposal requires review
+- whether proposal may auto-apply
+- whether post-apply verification passed
+
+
+### 6. Override Store
+
+Stores runtime-only configuration changes.
+
+This is where endpoint repairs should live.
+
+
+## Relation To `aiprovider`
+
+`aiprovider` should remain the model gateway.
+
+It should not become the full agent runtime.
+
+Recommended split:
+
+- `aiprovider`
+  - provider adaptation
+  - prompt transport
+  - model execution
+  - protocol compatibility
+
+- agent runtime
+  - orchestration
+  - signal handling
+  - tool selection
+  - proposal generation
+  - policy and audit
+
+This keeps provider concerns and agent behavior concerns separate.
+
+
+## Relation To Datasource Health
+
+Datasource health becomes one vertical slice of this architecture.
+
+Mapping:
+
+- signal:
+  - endpoint unreachable
+  - schema mismatch
+  - bad content type
+- finding:
+  - `failed`
+  - `schema_changed`
+  - `moved_endpoint_suspected`
+- proposal:
+  - runtime override suggestion
+- assessment:
+  - datasource health summary for operators
+
+
+## Relation To Situational Awareness
+
+Future situational-awareness capabilities should reuse the same flow:
+
+- raw telemetry becomes signals
+- anomaly detection becomes findings
+- LLM correlation becomes reasoning
+- operator-facing output becomes assessments
+- policy-approved mitigations become actions
+
+This lets the platform evolve from operational health governance into broader cyber/network posture workflows without changing the architecture.
+
+
+## Suggested Delivery Sequence
+
+### Phase A
+
+- finalize shared object model
+- implement health-oriented signal and finding storage
+
+### Phase B
+
+- implement Health Agent
+- generate proposals only
+
+### Phase C
+
+- implement Assessment Agent
+- expose structured assessments via API
+
+### Phase D
+
+- implement Correlation Agent
+- support multi-source incident grouping
+
+### Phase E
+
+- implement Recovery Agent with policy-gated runtime actions
+
+
+## Recommended First Build
+
+The first build should not try to implement every agent role.
+
+Recommended initial slice:
+
+- shared object model
+- health signals
+- health findings
+- Health Agent
+- proposal generation only
+
+This gives immediate value while preserving the longer-term architecture.
+
+
+## Non-Goals For The First Iteration
+
+- repository YAML auto-rewrites
+- unrestricted autonomous action
+- full incident graph reasoning
+- automatic large-scale remediation
+- agent-owned configuration source of truth
+
+
+## Summary
+
+Planet should treat agents as a reusable runtime for evidence, reasoning, proposals, and assessments.
+
+The datasource health use case is the first practical entrypoint, but the architecture should already assume future situational-awareness expansion.
+
+The safest path is:
+
+- deterministic checks first
+- agent reasoning second
+- proposals before actions
+- runtime overrides instead of default mutation

docs/plans/agents-agent-runtime-roadmap.md

@@ -0,0 +1,346 @@
+# Agent Runtime Roadmap
+
+## Overview
+
+This document connects three existing planning threads into one implementation roadmap:
+
+- `aiprovider` as the model gateway
+- datasource health governance as the first practical agent use case
+- situational awareness as the broader long-term target
+
+Related documents:
+
+- [aiprovider](/home/ray/dev/linkong/planet/docs/technical/agents-aiprovider.md)
+- [datasource-health-plan](/home/ray/dev/linkong/planet/docs/plans/agents-datasource-health-plan.md)
+- [agent-architecture-plan](/home/ray/dev/linkong/planet/docs/plans/agents-agent-architecture-plan.md)
+
+
+## Big Picture
+
+Planet should evolve in layers:
+
+1. stable model gateway
+2. deterministic health and evidence collection
+3. agent runtime for reasoning and proposal generation
+4. situational-awareness assessments and controlled actions
+
+This prevents the system from collapsing into a single giant "AI feature" with unclear boundaries.
+
+
+## Architecture Overview
+
+```mermaid
+flowchart TD
+    U["Frontend / Backend APIs / Operators"] --> B["Planet Backend"]
+    B --> H["Datasource Health Services"]
+    B --> R["Agent Runtime"]
+    R --> P["aiprovider"]
+    P --> M["OpenAI / Anthropic / MiniMax / Ollama / Local Models"]
+
+    C["Collectors / Snapshots / Logs / Alerts / BGP Signals"] --> S["Signal Store"]
+    H --> S
+    S --> E["Evaluation Layer"]
+    E --> F["Findings"]
+    F --> R
+
+    W["Web Search / Page Fetch / Docs Fetch"] --> R
+    R --> PR["Proposals"]
+    R --> AS["Assessments"]
+
+    PR --> O["Runtime Overrides / Review Queue / Tasks"]
+    AS --> SA["Situational Awareness APIs / UI"]
+
+    O --> V["Verification Loop"]
+    V --> S
+```
+
+
+## Role Boundaries
+
+### `aiprovider`
+
+Responsibilities:
+
+- provider compatibility
+- protocol adaptation
+- auth and model transport
+- request/response normalization
+
+Not responsible for:
+
+- agent orchestration
+- business workflows
+- datasource repair policy
+- situational-awareness domain logic
+
+
+### Backend
+
+Responsibilities:
+
+- stable business APIs
+- auth and permissions
+- task orchestration
+- health records
+- proposal and override persistence
+- assessment exposure
+
+
+### Agent Runtime
+
+Responsibilities:
+
+- consume findings and context
+- invoke LLMs via `aiprovider`
+- invoke tools such as web search
+- create proposals
+- create assessments
+- route to policy-controlled action paths
+
+
+## Delivery Sequence
+
+## Stage 1: Gateway Foundation
+
+Status:
+
+- already in place
+
+Delivered by current work:
+
+- `aiprovider`
+- multi-provider compatibility
+- backend AI facade
+- MiniMax / Anthropic-compatible support
+- request-id propagation
+
+Primary outcome:
+
+- the system already has a stable way to call models
+
+
+## Stage 2: Datasource Health MVP
+
+Goal:
+
+- establish deterministic health observability
+
+Key work:
+
+- health check task runner
+- health result table
+- datasource health APIs
+- UI visibility
+- collector endpoint override precedence cleanup
+
+Primary outcome:
+
+- Planet knows which collectors are healthy before asking an LLM anything
+
+
+## Stage 3: Health Agent
+
+Goal:
+
+- let the first agent role operate on health failures
+
+Key work:
+
+- convert health failures into signals/findings
+- invoke agent only for failed or suspicious cases
+- produce repair proposals with evidence and confidence
+
+Primary outcome:
+
+- Planet can suggest endpoint repairs without mutating defaults
+
+
+## Stage 4: Runtime Repair Application
+
+Goal:
+
+- safely apply approved datasource repair proposals
+
+Key work:
+
+- override storage
+- policy-gated apply flow
+- verification after apply
+- rollback path
+
+Primary outcome:
+
+- datasource repair becomes operationally useful without polluting repository defaults
+
+
+## Stage 5: Situational Awareness Assessments
+
+Goal:
+
+- reuse the same runtime for broader operator-facing assessment
+
+Key work:
+
+- normalize telemetry and incident evidence into signals/findings
+- build Assessment Agent
+- expose structured assessments through backend APIs and UI
+
+Primary outcome:
+
+- LLM output becomes evidence-backed situational summary, not just ad hoc chat output
+
+
+## Stage 6: Correlation and Controlled Actions
+
+Goal:
+
+- connect multiple sources into higher-level posture and event groupings
+
+Key work:
+
+- event correlation
+- incident grouping
+- recommendation scoring
+- controlled action routing
+
+Primary outcome:
+
+- Planet becomes a true agent-assisted situational-awareness system
+
+
+## Implementation Tracks
+
+These tracks can progress in parallel, but they should stay loosely coupled.
+
+### Track A: Config and Runtime Resolution
+
+Scope:
+
+- datasource defaults
+- overrides
+- runtime precedence
+- audit trails
+
+First milestone:
+
+- health-safe override layer
+
+
+### Track B: Health and Evidence
+
+Scope:
+
+- deterministic checks
+- failure categorization
+- signal and finding persistence
+
+First milestone:
+
+- datasource health record system
+
+
+### Track C: Agent Runtime
+
+Scope:
+
+- shared object model
+- orchestration flow
+- prompt/tool pipeline
+- policy integration
+
+First milestone:
+
+- Health Agent proposal pipeline
+
+
+### Track D: Situational Awareness
+
+Scope:
+
+- assessment schema
+- multi-source context assembly
+- operator-facing outputs
+
+First milestone:
+
+- structured assessment API
+
+
+## Shared Artifacts
+
+To avoid fragmentation, these artifacts should be shared across all future agent work.
+
+### Shared object model
+
+- `Signal`
+- `Finding`
+- `Proposal`
+- `Assessment`
+
+### Shared orchestration flow
+
+- collect
+- validate
+- classify
+- reason
+- propose or assess
+- review or apply
+- verify
+- archive
+
+### Shared policy model
+
+- read-only
+- propose-only
+- apply-limited
+
+
+## Recommended Next Concrete Steps
+
+1. Build Stage 2 first
+
+- datasource health records
+- deterministic checks
+- no automatic repair
+
+2. Then build Stage 3
+
+- Health Agent
+- proposal generation only
+
+3. Then Stage 4
+
+- override apply flow
+- rollback and verification
+
+4. Only after that start Stage 5
+
+- broader situational-awareness assessment workflows
+
+
+## Why This Order
+
+Because situational-awareness quality depends on reliable upstream data.
+
+If datasource health is weak:
+
+- agent reasoning quality will degrade
+- false explanations will increase
+- assessment trust will drop
+
+So datasource health is not a side task.
+
+It is the first operational foundation for the later situational-awareness system.
+
+
+## Summary
+
+Planet should be built as:
+
+- `aiprovider` for model access
+- backend services for orchestration and persistence
+- datasource health as the first evidence-governance layer
+- agent runtime as the reusable reasoning core
+- situational awareness as the long-term application layer
+
+That path keeps the architecture coherent and lets each phase produce useful functionality without forcing a rewrite later.

docs/plans/agents-datasource-health-plan.md

@@ -0,0 +1,486 @@
+# Datasource Health Plan
+
+## Overview
+
+This document defines a phased plan for datasource health governance.
+
+The goal is to make collectors observable, diagnosable, and recoverable when upstream APIs change, while avoiding unsafe automatic mutation of repository defaults.
+
+The key principle is:
+
+- do not let runtime automation rewrite repository default config
+
+Instead, split responsibilities across:
+
+- default config
+- runtime overrides
+- health check records
+- agent-generated repair proposals
+
+
+## Problem Statement
+
+Collectors currently depend on third-party APIs, data downloads, mirrored JSON files, archive links, and web pages.
+
+These upstream dependencies can fail in several ways:
+
+- endpoint becomes unreachable
+- endpoint still responds but schema changes
+- content-type changes
+- website shuts down or moves
+- mirror link disappears
+- HTML structure changes and scraping fails
+- endpoint requires a new path or new host
+
+We want a system that can:
+
+- detect datasource health degradation early
+- identify likely cause
+- search for updated endpoints when reasonable
+- apply safe runtime fixes without polluting default repo config
+- preserve auditability and rollback
+
+
+## Design Principles
+
+1. Default config is stable
+
+- `backend/app/core/data_sources.yaml` remains the repository baseline.
+- It should be changed intentionally through normal development flow, not by autonomous runtime agents.
+
+2. Runtime fixes are isolated
+
+- Emergency or adaptive fixes should live in a runtime override layer.
+- Overrides should be reversible and auditable.
+
+3. Deterministic checks come first
+
+- Use normal programmatic health checks before using LLMs.
+- Only call an agent when deterministic checks indicate a meaningful failure.
+
+4. Agents suggest before they mutate
+
+- Agents should produce proposals with evidence and confidence.
+- Application of a proposal should be controlled by policy.
+
+5. Every repair is attributable
+
+- Store what changed, why, who or what suggested it, and when it was applied.
+
+
+## Configuration Layers
+
+Recommended runtime precedence:
+
+1. datasource endpoint override
+2. datasource DB endpoint override
+3. repository default YAML
+4. collector internal fallback logic
+
+Definitions:
+
+- repository default YAML:
+  - `backend/app/core/data_sources.yaml`
+  - versioned baseline
+- datasource DB endpoint override:
+  - existing `DataSourceConfig.endpoint`
+  - current runtime override entrypoint
+- datasource endpoint override:
+  - a dedicated new override table
+  - used for health-repair and proposal application
+- collector internal fallback logic:
+  - final defensive fallback
+  - should be minimized over time
+
+
+## Recommended Architecture
+
+### 1. Deterministic Health Checks
+
+Each collector gets a health profile with checks such as:
+
+- endpoint resolves
+- HTTP request succeeds
+- status code is acceptable
+- content-type is expected
+- body parses successfully
+- minimum structural fields exist
+- sample item count is plausible
+- latency is within threshold
+
+Output states:
+
+- `healthy`
+- `degraded`
+- `failed`
+- `schema_changed`
+- `rate_limited`
+- `auth_required`
+
+
+### 2. Agent-Assisted Repair Discovery
+
+Only triggered when deterministic health checks fail or return suspicious structure.
+
+Agent responsibilities:
+
+- search for current official endpoint or replacement path
+- inspect likely upstream documentation or landing pages
+- compare candidate endpoint output to collector expectations
+- produce a repair proposal with confidence and evidence
+
+Agent should not directly modify repository defaults.
+
+
+### 3. Safe Runtime Repair Application
+
+Repair proposals can be:
+
+- reviewed manually
+- auto-applied only under strict low-risk policy
+
+Auto-apply should be limited to cases like:
+
+- same trusted domain
+- highly similar response structure
+- repeated successful verification
+- confidence above threshold
+
+
+## Phased Delivery Plan
+
+## Phase 1: Deterministic Health MVP
+
+Goal:
+
+- build health observability without automated repair
+
+Scope:
+
+- datasource health check task runner
+- datasource health result persistence
+- endpoint reachability + parse checks
+- dashboard or API visibility into health status
+
+Deliverables:
+
+- health check service
+- health check record table
+- status endpoint
+- scheduled or manual check trigger
+
+No agent usage yet.
+
+
+## Phase 2: Agent Repair Proposals
+
+Goal:
+
+- let agent investigate failing sources and propose updated endpoints
+
+Scope:
+
+- invoke agent only when datasource health is `failed` or `schema_changed`
+- web search + page inspection
+- candidate endpoint extraction
+- proposal persistence
+
+Deliverables:
+
+- repair proposal schema
+- proposal generation pipeline
+- confidence and evidence model
+- operator review view or API
+
+Still no automatic config mutation.
+
+
+## Phase 3: Runtime Overrides
+
+Goal:
+
+- allow approved proposals to take effect safely at runtime
+
+Scope:
+
+- add dedicated override storage
+- runtime resolution prefers override over default config
+- proposal application writes override only
+
+Deliverables:
+
+- endpoint override table
+- override-aware resolution logic
+- apply/reject endpoints
+- rollback endpoint
+
+Repository default YAML remains untouched.
+
+
+## Phase 4: Limited Auto-Apply
+
+Goal:
+
+- safely automate a narrow slice of low-risk repairs
+
+Scope:
+
+- policy engine for auto-apply
+- same-domain or trusted-domain checks
+- structure validation
+- staged verification after apply
+
+Deliverables:
+
+- auto-apply rules
+- audit logs
+- automatic post-apply health verification
+- auto-disable or rollback on regression
+
+
+## Data Model Draft
+
+### datasource_health_checks
+
+Purpose:
+
+- store each health evaluation result
+
+Suggested fields:
+
+- `id`
+- `datasource_id`
+- `collector_name`
+- `endpoint_checked`
+- `status`
+- `http_status`
+- `content_type`
+- `latency_ms`
+- `sample_count`
+- `error_message`
+- `details`
+- `checked_at`
+
+`details` can store structured diagnostic data such as:
+
+- parsed fields
+- schema mismatch summary
+- retry count
+- exception class
+
+
+### datasource_repair_proposals
+
+Purpose:
+
+- store agent-generated repair suggestions
+
+Suggested fields:
+
+- `id`
+- `datasource_id`
+- `collector_name`
+- `old_endpoint`
+- `candidate_endpoint`
+- `reason`
+- `confidence`
+- `evidence_urls`
+- `evidence_summary`
+- `status`
+- `created_by`
+- `created_at`
+- `reviewed_at`
+
+Suggested `status` values:
+
+- `proposed`
+- `approved`
+- `rejected`
+- `applied`
+- `expired`
+
+
+### datasource_endpoint_overrides
+
+Purpose:
+
+- runtime endpoint override layer
+
+Suggested fields:
+
+- `id`
+- `datasource_id`
+- `collector_name`
+- `endpoint`
+- `reason`
+- `source`
+- `proposal_id`
+- `enabled`
+- `created_at`
+- `updated_at`
+
+Suggested `source` values:
+
+- `manual`
+- `health-agent`
+- `migration`
+
+
+## API Draft
+
+### Health
+
+- `GET /api/v1/datasources/health`
+- `GET /api/v1/datasources/{id}/health`
+- `POST /api/v1/datasources/{id}/health-check`
+- `POST /api/v1/datasources/health-check-all`
+
+### Repair proposals
+
+- `GET /api/v1/datasources/{id}/repair-proposals`
+- `POST /api/v1/datasources/{id}/repair-proposals/generate`
+- `POST /api/v1/datasources/{id}/repair-proposals/{proposal_id}/approve`
+- `POST /api/v1/datasources/{id}/repair-proposals/{proposal_id}/reject`
+- `POST /api/v1/datasources/{id}/repair-proposals/{proposal_id}/apply`
+
+### Overrides
+
+- `GET /api/v1/datasources/{id}/overrides`
+- `POST /api/v1/datasources/{id}/overrides`
+- `PUT /api/v1/datasources/{id}/overrides/{override_id}`
+- `DELETE /api/v1/datasources/{id}/overrides/{override_id}`
+
+
+## Agent Contract Draft
+
+When deterministic health fails, the agent should receive:
+
+- datasource name
+- collector name
+- current endpoint
+- current failure mode
+- expected response shape summary
+- known trusted domains
+
+Expected output:
+
+```json
+{
+  "status": "proposal",
+  "candidate_endpoint": "https://example.com/api/v2/data",
+  "confidence": 0.86,
+  "reason": "Official docs now point to v2 endpoint",
+  "evidence_urls": [
+    "https://example.com/docs/api",
+    "https://example.com/changelog"
+  ],
+  "notes": "Response shape appears compatible after light field remapping"
+}
+```
+
+The agent should never output "rewrite the default yaml" as its primary action.
+
+
+## Risk Analysis
+
+### Risk: wrong endpoint chosen by agent
+
+Mitigation:
+
+- use trusted-domain allowlists
+- require evidence URLs
+- require confidence threshold
+- add manual review for medium-risk sources
+
+
+### Risk: endpoint responds but schema silently changed
+
+Mitigation:
+
+- deterministic schema checks
+- parse and sample validation
+- content-type checks
+- collector-specific required fields
+
+
+### Risk: automatic runtime override causes hidden drift
+
+Mitigation:
+
+- store all overrides explicitly
+- mark source of override
+- keep default YAML unchanged
+- expose active overrides in API/UI
+
+
+### Risk: persistent bad override breaks data collection
+
+Mitigation:
+
+- allow rollback
+- keep parent/default endpoint visible
+- re-run verification after apply
+- auto-disable override on repeated failure
+
+
+## Operational Policy Recommendations
+
+1. Do not auto-apply for high-value or high-fragility sources initially.
+
+2. Use manual approval for:
+
+- scraped HTML sources
+- unofficial mirrors
+- sources with auth or rate-limit complexity
+- sources with legal or trust ambiguity
+
+3. Allow auto-apply only for:
+
+- same-domain version bumps
+- obvious official migration paths
+- repeated passing verification
+
+4. Expose health + proposal + override state together in one operator view.
+
+
+## Suggested Implementation Order
+
+1. Phase 1
+- health result table
+- deterministic checks
+- API and UI visibility
+
+2. Phase 2
+- proposal table
+- agent prompt/output contract
+- proposal generation job
+
+3. Phase 3
+- runtime override table
+- resolver precedence update
+- apply/reject endpoints
+
+4. Phase 4
+- auto-apply rules
+- rollback policy
+- operator automation
+
+
+## Out Of Scope For The First Iteration
+
+- direct automatic mutation of repository default YAML
+- automatic git commits by repair agents
+- unrestricted autonomous endpoint replacement
+- fully generalized schema remapping engine
+
+
+## Recommended First Milestone
+
+The first milestone should be:
+
+- deterministic datasource health checks
+- persisted results
+- manual visibility
+- no automatic repair
+
+This gives immediate operational value with low risk, and prepares clean inputs for the later agent phase.

docs/plans/agents-datasource-health-stage2-tasks.md

@@ -0,0 +1,478 @@
+# Datasource Health Stage 2 Tasks
+
+## Goal
+
+Stage 2 focuses on the first practical operational layer:
+
+- deterministic datasource health checks
+- persisted health results
+- health visibility through API and UI
+- no agent-assisted repair yet
+
+This stage should make Planet capable of answering:
+
+- which collectors are healthy
+- which collectors are degraded
+- which collectors are failing
+- why they are failing at a basic deterministic level
+
+
+## Scope
+
+Included:
+
+- datasource health data model
+- deterministic health check service
+- manual and scheduled health check triggers
+- health result APIs
+- frontend visibility
+
+Excluded:
+
+- LLM reasoning
+- web-search-based repair proposals
+- automatic endpoint rewriting
+- runtime override application
+
+
+## Delivery Target
+
+At the end of Stage 2, an operator should be able to:
+
+1. see health status for each collector
+2. trigger a health check manually
+3. inspect the latest failure reason
+4. inspect the last checked endpoint
+5. understand whether the problem is:
+   - unreachable
+   - auth-related
+   - rate-limit-related
+   - schema-related
+   - empty-data-related
+
+
+## Work Breakdown
+
+## A. Data Model
+
+### A1. Add datasource health record table
+
+Create a new model, for example:
+
+- `backend/app/models/datasource_health_check.py`
+
+Suggested fields:
+
+- `id`
+- `datasource_id`
+- `collector_name`
+- `endpoint_checked`
+- `status`
+- `http_status`
+- `content_type`
+- `latency_ms`
+- `sample_count`
+- `error_message`
+- `details`
+- `checked_at`
+
+Suggested status enum values:
+
+- `healthy`
+- `degraded`
+- `failed`
+- `schema_changed`
+- `rate_limited`
+- `auth_required`
+- `empty_result`
+
+
+### A2. Add datasource health summary fields
+
+Option A:
+
+- keep summary only in the health check table
+
+Option B:
+
+- also add summary fields on `data_sources`
+
+Recommended first step:
+
+- do not mutate `data_sources` schema yet
+- derive summary from the latest health record
+
+
+### A3. Migration task
+
+Add migration for the health table.
+
+Deliverables:
+
+- migration file
+- model registration
+
+
+## B. Health Check Engine
+
+### B1. Define health check service
+
+Add a new service module, for example:
+
+- `backend/app/services/datasource_health.py`
+
+Responsibilities:
+
+- resolve effective endpoint
+- execute deterministic check
+- classify result
+- persist health record
+
+
+### B2. Define shared result schema
+
+Create a typed result object, for example:
+
+- `HealthCheckResult`
+
+Suggested fields:
+
+- `status`
+- `endpoint_checked`
+- `http_status`
+- `content_type`
+- `latency_ms`
+- `sample_count`
+- `error_message`
+- `details`
+
+
+### B3. Implement base deterministic checks
+
+Every datasource should go through a minimal baseline check:
+
+1. resolve endpoint
+2. perform request
+3. measure latency
+4. inspect status code
+5. inspect content type
+6. inspect body shape
+
+Classification rules:
+
+- network error -> `failed`
+- HTTP 401/403 -> `auth_required`
+- HTTP 429 -> `rate_limited`
+- HTTP 404/410 -> `failed`
+- parse failure -> `schema_changed`
+- zero or suspiciously empty results -> `empty_result` or `degraded`
+- valid parse -> `healthy`
+
+
+### B4. Add collector-aware adapters
+
+Some collectors do not use the same fetch semantics.
+
+Add adapter profiles such as:
+
+- `http_json`
+- `http_csv`
+- `html_scrape`
+- `stream_probe`
+- `auth_session_http`
+
+Initial mapping suggestion:
+
+- `huggingface`, `peeringdb`, `cloudflare` -> `http_json`
+- `fao` -> `http_csv`
+- `top500`, `epoch_ai`, `telegeography live_map` -> `html_scrape`
+- `ris_live` -> `stream_probe`
+- `spacetrack` -> `auth_session_http`
+
+
+### B5. Add sample validation hooks
+
+For each adapter, add a lightweight validation rule.
+
+Examples:
+
+- JSON array length > 0
+- CSV rows > 1
+- HTML page contains expected table or script patterns
+- stream source yields at least one valid event within timeout
+
+
+## C. Persistence and Query Layer
+
+### C1. Save every check run
+
+Each health check should insert a record.
+
+Do not overwrite history in Stage 2.
+
+
+### C2. Add latest-health query helpers
+
+Add helper functions to fetch:
+
+- latest health record by datasource
+- latest failed health record
+- recent health history
+
+
+### C3. Optional retention policy
+
+For Stage 2, retention can be deferred.
+
+If desired, keep only:
+
+- last N records per datasource
+
+
+## D. API Layer
+
+### D1. Add health list endpoint
+
+Suggested endpoint:
+
+- `GET /api/v1/datasources/health`
+
+Returns:
+
+- datasource id
+- collector name
+- current endpoint
+- latest health status
+- last checked time
+- short reason
+
+
+### D2. Add per-datasource health detail endpoint
+
+Suggested endpoint:
+
+- `GET /api/v1/datasources/{id}/health`
+
+Returns:
+
+- latest record
+- recent history
+- detailed classification fields
+
+
+### D3. Add manual health trigger endpoint
+
+Suggested endpoint:
+
+- `POST /api/v1/datasources/{id}/health-check`
+
+Behavior:
+
+- run a health check now
+- persist the result
+- return the new record
+
+
+### D4. Add bulk health trigger endpoint
+
+Suggested endpoint:
+
+- `POST /api/v1/datasources/health-check-all`
+
+Behavior:
+
+- enqueue or run health checks for all active datasources
+
+
+## E. Scheduling
+
+### E1. Add health scheduler task
+
+Decide scheduling strategy.
+
+Recommended first version:
+
+- run collector jobs and health checks separately
+- health checks run on a lower frequency
+
+Suggested frequency:
+
+- every 6h or 12h for most datasources
+- optionally on-demand only in the very first cut
+
+
+### E2. Prevent health check collision with collection
+
+Rules:
+
+- health checks should not disrupt active collection
+- they should use light requests
+- if a collector is currently running, health check may:
+  - skip
+  - or use a lightweight endpoint probe only
+
+
+## F. Frontend
+
+### F1. Add health columns to datasource list
+
+Update:
+
+- `frontend/src/pages/DataSources/DataSources.tsx`
+
+Suggested new columns:
+
+- health status
+- last checked
+- reason summary
+
+
+### F2. Add manual health check action
+
+Per datasource:
+
+- button or dropdown action:
+  - `健康检查`
+
+
+### F3. Add health detail drawer or modal
+
+Show:
+
+- endpoint checked
+- status
+- HTTP status
+- content type
+- sample count
+- error message
+- last few results
+
+
+### F4. Add basic visual language
+
+Suggested colors:
+
+- green -> healthy
+- yellow -> degraded
+- orange -> rate-limited / auth-required
+- red -> failed / schema-changed
+
+
+## G. Observability
+
+### G1. Structured logging
+
+Every health check should log:
+
+- datasource id
+- collector name
+- endpoint
+- status
+- latency
+- failure class
+
+
+### G2. Optional metrics
+
+If metrics are added later, useful counters include:
+
+- health checks total
+- health checks failed
+- schema changes detected
+- rate limited checks
+
+
+## H. Tests
+
+### H1. Unit tests
+
+Add tests for:
+
+- status classification
+- content type classification
+- adapter behavior
+- latest-health query helpers
+
+
+### H2. API tests
+
+Add tests for:
+
+- health endpoints require auth
+- manual trigger endpoint works
+- list endpoint returns latest status
+
+
+### H3. Failure-path tests
+
+Add coverage for:
+
+- HTTP 404
+- HTTP 429
+- invalid JSON
+- empty response
+- parse mismatch
+
+
+## Suggested File Plan
+
+Possible implementation files:
+
+- `backend/app/models/datasource_health_check.py`
+- `backend/app/services/datasource_health.py`
+- `backend/app/schemas/datasource_health.py`
+- `backend/app/api/v1/datasource_health.py`
+- migration file under the project migration system
+
+Likely touched existing files:
+
+- `backend/app/api/main.py`
+- `frontend/src/pages/DataSources/DataSources.tsx`
+- `backend/tests/test_api.py`
+
+
+## Suggested Execution Order
+
+1. Add model and migration
+2. Add service and result schema
+3. Add deterministic adapters
+4. Add manual trigger API
+5. Add list/detail API
+6. Add frontend visibility
+7. Add scheduled checks
+8. Expand tests
+
+
+## Minimal First Milestone
+
+If we want the fastest useful slice, do this first:
+
+1. health table
+2. deterministic check service
+3. manual per-datasource health check API
+4. latest health list API
+5. frontend status badge column
+
+That is enough to start operating the system and will provide the input layer for Stage 3.
+
+
+## Dependency On Later Stages
+
+Stage 2 outputs become direct inputs for Stage 3.
+
+Specifically:
+
+- failed or schema-changed health records become agent triggers
+- health history becomes repair context
+- endpoint_checked becomes proposal baseline
+
+
+## Success Criteria
+
+Stage 2 is done when:
+
+- every active datasource can be health-checked deterministically
+- the latest health state is visible in API and UI
+- operators can manually trigger checks
+- failures are categorized into stable machine-readable statuses
+- no LLM is required for core health visibility

docs/plans/agents-situational-awareness-foundation-plan.md

@@ -0,0 +1,309 @@
+# Situational Awareness Foundation Plan
+
+## 定位
+
+当前这套 AI 能力应被视为 `态势感知服务底座`，而不是完整的态势感知产品。
+
+也就是说，现阶段的目标不是：
+
+- 做一个“什么都能分析”的万能 AI 页面
+- 让模型在证据不足时替代人工研判
+- 过早把页面做成完整指挥大屏
+
+现阶段真正要做的是：
+
+- 先把 `model gateway / backend facade / evidence injection / page-specific brief` 这几层边界搭稳
+- 让系统能够在已有证据上稳定地产出“可读、可回看、可扩展”的摘要
+- 为后续更强的数据联动、agent 推理和 assessment 结构化输出预留好接口与数据模型
+
+## 当前现实约束
+
+### 1. 数据维度不足
+
+目前系统能提供的主要证据仍集中在：
+
+- BGP incidents / anomalies / events
+- collector coverage
+- datasource health / platform alerts
+- prefix geography 的部分归属信息
+
+当前明显还缺：
+
+- 流量异常与业务指标
+- 电商、支付、物流等业务侧指标
+- 更丰富的资产、链路、区域、行业画像
+- 外部舆情、公告、运营商状态、基础设施事件等背景信息
+
+这意味着：
+
+- 模型现在可以做“基于现有证据的摘要与归纳”
+- 但还不能可靠地做“跨维度因果研判”
+
+### 2. 维度之间联动还弱
+
+目前不同模块之间更多是“并列展示”，还不是“强关联分析”：
+
+- 系统告警和 BGP 事件还没有统一事件模型
+- collector bias 与真实区域热度还没有完全剥离
+- datasource health 与 BGP 风险、业务影响之间还没有稳定映射
+
+这意味着：
+
+- 当前更适合做 `brief / overview / operator notes`
+- 还不适合过度承诺“自动态势判断”
+
+### 3. 结构化 assessment 还未成为主输出
+
+虽然已经有 BGP brief、系统告警 brief、态势告警 brief，但目前主输出仍偏向：
+
+- 文本摘要
+- facts/context 附带证据
+
+后续真正要服务态势感知，需要更稳定的结构化输出，例如：
+
+- summary
+- key risks
+- evidence
+- confidence
+- recommendations
+- missing data
+
+## 当前基座已经具备的能力
+
+### 1. AI 调用边界已经明确
+
+- `aiprovider` 负责模型协议与 provider 兼容
+- `backend` 负责业务 API、证据整合和鉴权
+- `frontend` 负责页面入口与结果展示
+
+### 2. 页面级 AI 入口已经开始成型
+
+当前已经有或正在收口的入口：
+
+- `Playground`
+  - 用于链路验证与 provider 诊断
+- `BGP AI 简报`
+  - 用于 BGP 事实摘要和区域风险归纳
+- `Alerts`
+  - 用于系统告警、BGP 告警、态势告警三类入口
+
+### 3. 证据优先的方向已经建立
+
+已经不再只依赖人工在 Playground 中手填 prompt，系统开始具备：
+
+- 从真实业务数据生成事实输入
+- 保存 facts/context 快照
+- 回看 AI 输出时同时回看证据
+
+这一步非常关键，因为它决定后面能否从“玩具 demo”走向“有运维价值的系统”。
+
+## 近期收尾建议
+
+这些事情都属于“底座收口”，值得做，但不应该再继续重产品包装。
+
+### 1. 统一 Alerts 页面
+
+已采用：
+
+- 一个 `Alerts` 页面
+- 三个 tab：
+  - `系统告警`
+  - `BGP 告警`
+  - `态势告警`
+
+收尾重点：
+
+- 保持 tab 的文案、摘要卡和 AI 简报交互一致
+- 不额外扩展成多个独立二级页面
+
+### 2. 保持 Playground 为测试台
+
+原则：
+
+- Playground 只承担链路验证、provider 状态诊断、请求结果观察
+- 不继续堆“万能业务分析器”式交互
+
+### 3. 把 brief 能力当服务能力而不是页面特效
+
+页面现在能看到按钮和结果，这很好，但更重要的是：
+
+- 后端接口稳定
+- facts/context 可追踪
+- 输出结构后续可升级
+
+### 4. 导航结构先收口，不继续平铺一级菜单
+
+随着后续能力扩展，系统很可能继续新增：
+
+- 海缆
+- 算力中心
+- 战争信息
+- 电商分析
+- 其他专题观测页
+
+如果继续把这些入口全部平铺在左侧一级菜单中，会带来两个问题：
+
+- 一级菜单过长，用户难以判断先进入哪个上下文
+- `观测页 / 告警页 / 研判页 / 运维页` 的职责边界会被混在一起
+
+因此近期应明确采用分组导航，而不是继续扩展平铺菜单。
+
+推荐的导航分组如下：
+
+- `总览`
+  - 仪表盘
+  - Earth
+- `专题观测`
+  - BGP 观测
+  - 采集数据
+  - 后续可扩展：海缆、算力中心、战争信息、电商分析
+- `告警与研判`
+  - Alerts
+- `运维与配置`
+  - 数据源
+  - AI Playground
+  - 用户管理
+  - 系统配置
+
+这套结构的含义是：
+
+- `专题观测` 页面负责看某个维度本身
+- `Alerts` 负责跨模块风险与值班工作台
+- `Playground` 保持为测试台，不挤占业务导航语义
+
+短期收尾时，应优先重组现有入口，而不是继续增加新的一级菜单。
+
+## 后续路线
+
+## Phase 1：服务底座稳固
+
+目标：
+
+- 不追求“更炫的 AI 页面”
+- 先把当前接口、证据、存储和页面入口收稳
+
+工作项：
+
+- 统一页面级 AI 入口模式
+- 统一 brief response schema
+- 保证 facts/context 在前后端都可回看
+- 继续清理 mock 和临时分支逻辑
+
+完成标准：
+
+- 每个 AI 入口都是真实链路
+- 每个 AI 结果都能追溯到证据输入
+
+## Phase 2：Evidence-first Assessment
+
+目标：
+
+- 从“文本摘要”升级成“结构化 assessment”
+
+工作项：
+
+- 为 brief/assessment 定义统一 schema
+- 固化：
+  - summary
+  - key_risks
+  - evidence
+  - confidence
+  - recommendations
+  - missing_data
+- 页面以结构化区块展示，而不只是大段文本
+
+完成标准：
+
+- AI 输出可持久化、可比较、可审计
+
+## Phase 3：多维证据接入
+
+目标：
+
+- 让“态势感知”真正拥有更多维度，而不是只靠 BGP 与系统告警
+
+优先接入方向：
+
+- datasource health findings
+- 流量或业务指标
+- 区域/资产/链路映射
+- 外部事件与公告
+- 业务垂直数据，例如电商分析相关指标
+
+完成标准：
+
+- AI 能基于多个维度做交叉说明
+- 不再只围绕单一模块自说自话
+
+## Phase 4：Correlation Layer
+
+目标：
+
+- 不同来源的信号不再只是并列，而是形成统一的事件关联
+
+工作项：
+
+- 统一 signal/finding 模型
+- 跨模块事件聚合
+- 证据来源权重
+- collector bias 与真实热度分离
+
+完成标准：
+
+- 系统能回答“这些异常是不是同一件事”
+- 系统能回答“哪些结论只是观测偏差”
+
+## Phase 5：Agent-assisted Situational Awareness
+
+目标：
+
+- 在证据足够的前提下，再让 agent 负责更复杂的推理与建议
+
+工作项：
+
+- 复用现有 agent runtime 规划
+- 引入 web search / docs fetch / repair proposal 等能力
+- 但始终坚持：
+  - evidence first
+  - proposal before action
+  - no silent mutation of defaults
+
+完成标准：
+
+- agent 成为证据驱动的分析层
+- 而不是一个“万能猜测层”
+
+## 设计原则
+
+### 1. 先底座，后产品化
+
+先把服务链路和证据模型做好，再做更大的页面表达。
+
+### 2. 先证据，后判断
+
+事实输入应先稳定，再让模型做归纳。
+
+### 3. 先专用 brief，后统一态势层
+
+先让各业务页有各自可信的 AI 入口，再考虑统一态势页。
+
+### 4. 先 proposal，后自动动作
+
+涉及修复、覆盖、写配置、调任务的动作，都应经过 proposal 和审计。
+
+## 当前建议结论
+
+对现在这个项目，最合理的定位是：
+
+- `Playground` 是测试台
+- `BGP / Alerts` 是第一批业务 AI 入口
+- `aiprovider + backend AI facade + evidence snapshots` 是核心服务底座
+
+现阶段不需要追求“已经具备完整态势感知能力”。
+
+现阶段真正的成功标准是：
+
+- 这套底座可用
+- 可回看
+- 可扩展
+- 不自欺欺人

docs/plans/earth-bgp-region-aggregation-plan.md

@@ -17,7 +17,7 @@ It is an aggregation/view-model layer:
 
 ## Why This Layer Exists
 
-Current product gap from [bgp-context.md](/home/ray/dev/linkong/planet/docs/bgp-context.md):
+Current product gap from [bgp-context.md](/home/ray/dev/linkong/planet/docs/technical/earth-bgp-context.md):
 
 - incident density is naturally low
 - anomaly density is higher, but still not enough to keep the globe expressive all the time
@@ -290,7 +290,7 @@ Each feature should include:
 
 ## Earth Rendering Plan
 
-Detailed visual layering guidance is expanded in [bgp-earth-rendering-plan.md](/home/ray/dev/linkong/planet/docs/bgp-earth-rendering-plan.md).
+Detailed visual layering guidance is expanded in [bgp-earth-rendering-plan.md](/home/ray/dev/linkong/planet/docs/plans/earth-bgp-earth-rendering-plan.md).
 
 ### Layer Relationship
 

docs/plans/earth-celestial-background-plan.md

@@ -0,0 +1,715 @@
+# Earth 天球背景与日月位置实施方案
+
+## 目标
+
+为 Earth 大屏增加一套真正可用的天文背景层，覆盖三件事：
+
+1. 用真实天球背景替换当前随机星点
+2. 在当前时间下显示太阳与月亮的相对位置
+3. 让太阳方向同时驱动地球受光，形成更可信的昼夜关系
+
+本方案优先追求：
+
+- 与当前 Three.js Earth 架构兼容
+- 风险可控
+- 先落地一版真实感明显提升的 V1
+- 为后续更严格的天文参考系升级预留余地
+
+## 当前现状
+
+当前 Earth 的基础条件已经具备：
+
+- 地球、云层、地形、网格都基于 Three.js，主渲染入口在 [frontend/public/earth/js/main.js](/home/ray/dev/linkong/planet/frontend/public/earth/js/main.js)
+- 地球实体创建在 [frontend/public/earth/js/earth.js](/home/ray/dev/linkong/planet/frontend/public/earth/js/earth.js)
+- 当前所谓“宇宙背景”只是 `createStars()` 生成的随机星点，不是真实星图
+- Earth 已有倾角常量 `EARTH_CONFIG.tiltRad`，位于 [frontend/public/earth/js/constants.js](/home/ray/dev/linkong/planet/frontend/public/earth/js/constants.js)
+- 主循环 `animate()` 已稳定运行，可在其中接入天体更新逻辑
+
+这意味着：
+
+- 不需要重写 Earth
+- 可以在现有 scene/world 层新增一个 celestial layer
+- 第一阶段不必拆 Earth / satellite / cable 的参考系
+
+## 总体策略
+
+采用“两层现实”设计：
+
+### 1. 世界层（world-space celestial layer）
+
+用于放置：
+
+- 天球背景
+- 太阳
+- 月亮
+- 太阳光方向
+
+这些对象不挂在 `earthObj` 上，而是直接放在 `scene` 中。
+
+### 2. 地球层（earth-fixed layer）
+
+继续保持当前结构：
+
+- 海缆
+- 登陆点
+- 卫星点与轨迹
+- BGP 覆盖
+- 地球纹理、云层、地形
+
+这些对象继续挂在 `earthObj` 下，不打断现有交互。
+
+## 为什么先这样做
+
+当前用户交互是“拖动地球本体”，而不是“移动相机绕惯性系观测”。  
+如果现在直接做严格惯性参考系改造，会同时影响：
+
+- `earthObj.rotation`
+- 卫星轨迹与锁定逻辑
+- 海缆与登陆点附着关系
+- resetView / autoRotate / hover / click 等交互链路
+
+所以第一阶段只做：
+
+- 真正的天空
+- 真正的日月方向
+- 不碰现有 Earth 附着对象的语义
+
+## 推荐技术选型
+
+### 天文计算库
+
+推荐：
+
+- [Astronomy Engine](https://github.com/cosinekitty/astronomy)
+
+原因：
+
+- 有 JavaScript 版本
+- 支持 Sun / Moon 的矢量与坐标变换
+- 精度、可扩展性都比轻量太阳高度角库更适合本项目
+- 后续若要加行星、月相、黄道、赤道网，也能继续沿用
+
+不作为主选的库：
+
+- [SunCalc](https://github.com/mourner/suncalc)
+
+原因：
+
+- 更偏本地观察者视角的太阳/月亮高度角
+- 用于“地面日出日落”很好
+- 但不如 Astronomy Engine 适合做真实天球与后续空间参考系扩展
+
+### Three.js 表现层
+
+推荐组合：
+
+- 天球：内翻球壳 + 星图纹理
+- 太阳：`THREE.Sprite`
+- 月亮：`THREE.Sprite` 或小型 `THREE.Mesh`
+- 太阳光：`THREE.DirectionalLight`
+
+参考：
+
+- [Three.js SpriteMaterial](https://threejs.org/docs/pages/SpriteMaterial.html)
+
+## 天球背景资源与星体数据来源
+
+为避免把“视觉背景”和“可计算天体位置”混为一谈，本方案明确分成两类资源：
+
+### 1. 背景资源：全天星图贴图
+
+用于 Phase 1 的“真实天空背景”。
+
+推荐优先来源：
+
+- NASA SVS 的 Tycho 全天星图
+  - [The Tycho Catalog Skymap - Version 2.0](https://svs.gsfc.nasa.gov/3572/)
+- NASA Deep Star Maps 2020
+  - SatelliteMap.space 在 credits 中明确提到其使用了 `NASA Deep Star Maps 2020 - High-resolution star field (1.7 billion stars from Gaia DR2)` 作为星空视觉资源
+  - 这说明行业内成熟实现并不一定直接渲染全部星表点，而很可能先使用一张高质量官方深空星图作为背景层
+- 如需后续替换，也可评估 ESA / Gaia 的全天 sky map 资源
+  - [Gaia DR3 stories](https://www.cosmos.esa.int/web/gaia/dr3-stories)
+
+建议要求：
+
+- 使用官方来源或官方衍生可复用资源
+- 等距矩形投影（equirectangular）
+- 坐标定义尽量明确为赤道坐标展开
+- 分辨率建议至少 `4k`
+- 颜色不要过亮，避免压过 Earth HUD 前景
+- 尽量优先选择官方天文机构已经生产好的深空图，而不是自行拼接低质量星空纹理
+
+建议本地资源目录：
+
+- `frontend/public/earth/assets/celestial/starmap_equatorial_4k.jpg`
+
+### 2. 位置数据：星表与天体计算
+
+用于 Phase 2+ 的“位置正确的星体”。
+
+推荐来源分两层：
+
+- 太阳、月亮位置
+  - 使用 [Astronomy Engine](https://github.com/cosinekitty/astronomy)
+- 恒星位置
+  - 第一优先：Hipparcos / Tycho
+    - [Hipparcos overview](https://www.cosmos.esa.int/web/Hipparcos)
+    - [Hipparcos catalogues](https://www.cosmos.esa.int/web/hipparcos/catalogues)
+  - 第二优先：Gaia
+    - [Gaia DR3 stories](https://www.cosmos.esa.int/web/gaia/dr3-stories)
+
+建议策略：
+
+- V1：背景球壳只用全天星图，不立即生成全量恒星点
+- V2：只挑选亮星（例如星等 `< 5.5`）生成恒星点层
+- V3：如果确实需要更丰富的星场，再逐步扩展到更深星等
+
+这样做的原因：
+
+- 背景球壳负责“天球真实感”
+- 亮星点负责“位置正确、可后续标注和高亮”
+- 不需要一开始就处理数十万甚至数百万颗星
+
+### 3. 对外部成熟实现的参考结论
+
+`SatelliteMap.space` 的公开 credits 提供了一个很有价值的参考样板：
+
+- 图形渲染使用 `TWGL.js`
+- 天文计算使用 `Skyfield` 与 `Astronomia`
+- 星空/天球视觉资源使用 `NASA Deep Star Maps 2020`
+
+这给本项目的启发是：
+
+- “真实感强的天球背景”完全可以先依赖官方高质量深空图
+- “位置正确的动态天体”则应依赖单独的天文计算链路
+- 没有必要在第一版就直接渲染完整星表
+
+因此本项目推荐继续坚持两层拆分：
+
+- 背景层：官方深空图 / 全天星图
+- 计算层：太阳、月亮与后续亮星点
+
+## 如何保证星体位置正确
+
+位置正确不是只看“图看起来像”，而是要统一参考系和转换链路。
+
+### 1. 统一坐标基准
+
+本方案推荐统一使用：
+
+- `J2000` 赤道坐标系作为恒星位置基准
+
+原因：
+
+- Hipparcos / Tycho 资料和大量天文可视化都容易映射到该基准
+- 太阳、月亮也可以通过 Astronomy Engine 转到同一坐标系
+- 这样背景、恒星点、太阳、月亮就能共用一套 sky orientation
+
+### 2. 背景贴图与点位必须使用同一展开逻辑
+
+如果背景球壳使用赤道坐标全天图，那么：
+
+- 亮星点也必须按赤道坐标贴到同一球面方向
+- 太阳/月亮 sprite 也必须按赤道坐标转换后落到同一 world-space
+
+否则会出现：
+
+- 背景银河带是对的
+- 但太阳/月亮或亮星点飘到不匹配的位置
+
+### 3. RA / Dec 到 Three.js 坐标的落点方式
+
+亮星点和日月方向最终都要转成单位球面向量。
+
+概念步骤：
+
+1. 读取赤经 `RA`
+2. 读取赤纬 `Dec`
+3. 转成弧度
+4. 映射到单位球面向量
+5. 再根据 Three.js 当前世界坐标定义做轴向映射
+
+参考公式：
+
+```text
+x = cos(dec) * cos(ra)
+y = sin(dec)
+z = cos(dec) * sin(ra)
+```
+
+实际接入 Three.js 时，需要做一次项目内坐标轴校准：
+
+- 验证 `RA = 0h`
+- 验证 `RA = 6h`
+- 验证北天极
+- 验证银河带主方向
+
+然后确定最终的：
+
+- `x/y/z` 对应 Three.js 哪个轴
+- 是否需要 `z` 取反
+- 是否需要整体再做一个固定 `rotation`
+
+建议把这层显式封装在：
+
+```js
+function equatorialToWorldVector(raRad, decRad)
+```
+
+不要把轴映射散落在不同模块里。
+
+### 4. 背景球壳与恒星点的关系
+
+推荐最终组合：
+
+- 背景层：全天星图球壳
+- 点位层：亮星点
+- 动态层：太阳 / 月亮
+
+这样有三个好处：
+
+- 背景层提供密集真实的天空纹理
+- 亮星点提供位置正确、可扩展的标注基础
+- 太阳/月亮提供与时间相关的真实动态对象
+
+## 数据与资源建议清单
+
+### 推荐首批引入资源
+
+1. 全天星图
+- 来源：NASA Tycho all-sky map
+- 用途：背景球壳纹理
+
+2. 月亮纹理
+- 用途：Phase 4 月相表现
+- 路径建议：
+  - `frontend/public/earth/assets/celestial/moon_albedo_2k.jpg`
+
+3. 太阳 glow 贴图
+- 用途：太阳 sprite halo
+- 路径建议：
+  - `frontend/public/earth/assets/celestial/sun_glow.png`
+
+### 推荐首批数据文件
+
+如果要上亮星层，建议新增一个预处理后的轻量数据文件：
+
+- `frontend/public/earth/assets/celestial/bright-stars.json`
+
+建议字段：
+
+```json
+[
+  {
+    "id": 32349,
+    "name": "Sirius",
+    "raDeg": 101.2875,
+    "decDeg": -16.7161,
+    "mag": -1.46,
+    "colorIndex": 0.00
+  }
+]
+```
+
+建议不要在浏览器里直接吞原始 Gaia 大表，而是先离线裁剪成：
+
+- 只保留亮星
+- 只保留渲染必需字段
+- JSON 或二进制轻量格式
+
+## 资源与数据实施路线
+
+### 路线 A：先做可用版本（推荐）
+
+1. 引入 NASA Tycho 全天图
+   - 或评估替换为更接近 SatelliteMap.space 路线的 `NASA Deep Star Maps 2020`
+2. 实现背景球壳
+3. 用 Astronomy Engine 计算太阳/月亮方向
+4. 暂不做亮星点
+
+优点：
+
+- 最快见效
+- 风险最低
+- 就能明显提升天球真实感
+
+### 路线 B：在 A 基础上增强
+
+1. 离线生成 `bright-stars.json`
+2. 浏览器端渲染亮星点
+3. 后续可加：
+   - 星座线
+   - 亮星名称
+   - 特定星体高亮
+
+优点：
+
+- 背景真实感和“位置正确的可交互星体”同时兼顾
+
+## 代码模块建议细化
+
+### 新增模块
+
+- `frontend/public/earth/js/celestial.js`
+  - 管理天球背景
+  - 管理太阳/月亮
+  - 管理亮星层（后续）
+
+- `frontend/public/earth/js/celestial-data.js`
+  - 资源路径
+  - 星图方向配置
+  - 亮星数据加载（后续）
+
+### 建议函数设计
+
+```js
+export function initCelestialLayer(scene)
+export function updateCelestialLayer(date)
+export function setCelestialVisibility(visible)
+export function disposeCelestialLayer()
+
+function loadStarMapTexture()
+function createSkySphere(texture)
+function createSunSprite()
+function createMoonSprite()
+function getSunEquatorialPosition(date)
+function getMoonEquatorialPosition(date)
+function equatorialToWorldVector(raRad, decRad)
+```
+
+### 推荐后续预处理脚本
+
+如要引入亮星层，建议单独做离线脚本：
+
+- `scripts/build_bright_stars.py`
+
+职责：
+
+- 从 Hipparcos / Tycho 源数据读取
+- 过滤亮星
+- 生成 `bright-stars.json`
+
+这样浏览器端只消费轻量结果，不承担大表解析成本。
+
+## 分阶段实施
+
+## Phase 1：真实天球背景
+
+### 目标
+
+用真实全天星图替换当前随机星点背景。
+
+### 做法
+
+1. 新增一张全天星图纹理
+
+建议路径：
+
+- `frontend/public/earth/assets/celestial/starmap_equatorial_4k.jpg`
+
+纹理要求：
+
+- 等距矩形投影
+- 赤经/赤纬坐标展开
+- 无地平线、无地景遮挡
+- 尽量深色、弱干扰，适合大屏 HUD 叠加
+
+2. 新增天球球壳
+
+新增模块：
+
+- [frontend/public/earth/js/celestial.js](/home/ray/dev/linkong/planet/frontend/public/earth/js/celestial.js)
+
+建议接口：
+
+```js
+export function initCelestialLayer(scene)
+export function updateCelestialLayer(date, camera, earth)
+export function disposeCelestialLayer()
+```
+
+3. 实现一个大半径内翻球体
+
+建议参数：
+
+- 半径：`600 ~ 900`
+- 材质：`MeshBasicMaterial`
+- `side: THREE.BackSide`
+- 不受场景光照影响
+- 始终围绕场景中心
+
+### 验收标准
+
+- 初始加载后背景不再是随机星点
+- 旋转地球时，背景保持为稳定天球而不是跟地球一起转
+- 不明显干扰海缆/卫星/BGP 的前景识别
+
+## Phase 2：太阳与月亮真实位置
+
+### 目标
+
+在当前 UTC 时间下，计算太阳与月亮在天球中的方向，并显示出来。
+
+### 做法
+
+1. 在 `celestial.js` 内封装天体位置计算
+
+建议函数：
+
+```js
+function getSunDirection(date)
+function getMoonDirection(date)
+```
+
+输出统一为 world-space `THREE.Vector3`
+
+2. 太阳显示
+
+- 一个暖色发光 sprite
+- 比月亮更大、更亮
+- 可选添加柔和 halo
+
+3. 月亮显示
+
+- 一个较小 sprite 或 sphere
+- 灰白偏冷色
+- 后续 Phase 3 再做月相
+
+4. 更新频率
+
+不要每帧重新做完整天文计算，建议：
+
+- 每 30 秒或 60 秒重算一次真实位置
+- 渲染帧内做平滑过渡
+
+### 验收标准
+
+- 页面可见太阳与月亮两个对象
+- 时间变化时位置会更新
+- 日月不会跟随地球局部旋转而错误附着
+
+## Phase 3：太阳驱动地球受光
+
+### 目标
+
+让地球光照方向与太阳方向一致，不再使用写死的固定主光。
+
+### 做法
+
+1. 替换或接管当前主定向光
+
+当前 [frontend/public/earth/js/main.js](/home/ray/dev/linkong/planet/frontend/public/earth/js/main.js) 中 `addLights()` 里用了固定方向的 `DirectionalLight`。
+
+建议改为：
+
+- 保留环境补光
+- 主太阳光方向由 `sunDirection` 决定
+
+2. 太阳光参数建议
+
+- `DirectionalLight` 颜色偏暖白
+- 强度略高于当前主光
+- 保留一个弱背光作为氛围补偿，避免背面过死黑
+
+3. 先不做物理级大气散射
+
+第一版只要求：
+
+- 亮面与暗面方向真实
+- 云层和大气仍保持当前风格
+
+### 验收标准
+
+- 地球明暗面会随太阳方向改变
+- 太阳 sprite 和地球亮面方向一致
+- 不破坏现有海缆、卫星、BGP 的可见性
+
+## Phase 4：月相与天文细节增强
+
+### 目标
+
+在日月真实位置基础上增加更强的“天文可信度”。
+
+### 可选项
+
+1. 月相
+
+- 根据日月夹角计算 illuminated fraction
+- 用月相纹理或 shader 表达盈亏
+
+2. 赤道/黄道辅助线
+
+- 可作为开发调试层，不默认显示
+
+3. 太阳 terminator 增强
+
+- 给地球夜面加入更自然的 night tint
+- 未来可叠加城市夜光纹理
+
+4. 天文时间入口
+
+- 设置中加入“当前时刻 / 指定时刻 / 加速时间”模式
+
+### 验收标准
+
+- 月亮不再只是一个静态圆点
+- 后续扩展行星或观测模式时无需推倒重来
+
+## Phase 5：严格参考系升级（可选，不作为 V1 必做）
+
+### 目标
+
+把 Earth 从“用户旋转球体”升级为“真实地球姿态 + 用户观察姿态”的双层模型。
+
+### 需要处理的问题
+
+- 地球自转角与 UTC 的一致性
+- 赤道坐标系、地固坐标系、相机交互层分离
+- 卫星轨道显示与 Earth 旋转同步关系
+- resetView 和 autoRotate 的语义重定
+
+### 风险
+
+这一步会影响：
+
+- [frontend/public/earth/js/main.js](/home/ray/dev/linkong/planet/frontend/public/earth/js/main.js)
+- [frontend/public/earth/js/satellites.js](/home/ray/dev/linkong/planet/frontend/public/earth/js/satellites.js)
+- [frontend/public/earth/js/cables.js](/home/ray/dev/linkong/planet/frontend/public/earth/js/cables.js)
+- [frontend/public/earth/js/controls.js](/home/ray/dev/linkong/planet/frontend/public/earth/js/controls.js)
+
+因此不建议与 V1 同时推进。
+
+## 代码改造清单
+
+## 1. 新增文件
+
+- [frontend/public/earth/js/celestial.js](/home/ray/dev/linkong/planet/frontend/public/earth/js/celestial.js)
+
+职责：
+
+- 管理天球背景、太阳、月亮
+- 对外暴露 init/update/dispose
+
+## 2. 修改 `constants.js`
+
+文件：
+
+- [frontend/public/earth/js/constants.js](/home/ray/dev/linkong/planet/frontend/public/earth/js/constants.js)
+
+新增：
+
+```js
+export const CELESTIAL_CONFIG = {
+  sphereRadius: 800,
+  updateIntervalMs: 60000,
+  sunSpriteScale: 28,
+  moonSpriteScale: 16,
+  sunLightIntensity: 1.25,
+  ambientIntensity: 0.28,
+  backLightIntensity: 0.18,
+};
+```
+
+## 3. 修改 `main.js`
+
+文件：
+
+- [frontend/public/earth/js/main.js](/home/ray/dev/linkong/planet/frontend/public/earth/js/main.js)
+
+主要改动：
+
+1. `init()` 中：
+   - 初始化 celestial layer
+2. `addLights()` 中：
+   - 把固定太阳光改成可更新的 celestial sun light
+3. `animate()` 中：
+   - 每帧调 `updateCelestialLayer()`
+4. `destroy()` 中：
+   - 清理 celestial 资源
+
+## 4. 修改 `earth.js`
+
+文件：
+
+- [frontend/public/earth/js/earth.js](/home/ray/dev/linkong/planet/frontend/public/earth/js/earth.js)
+
+主要改动：
+
+- `createStars()` 逐步退役
+- 第一阶段可先保留作为 fallback
+- 当真实星图加载成功后，不再显示随机星点
+
+## 5. 新增资源
+
+目录建议：
+
+- `frontend/public/earth/assets/celestial/`
+
+建议至少包含：
+
+- `starmap_equatorial_4k.jpg`
+- `sun_glow.png`
+- `moon_albedo_2k.jpg`
+
+## 数据流设计
+
+```mermaid
+flowchart TD
+    A["main.js:init()"] --> B["initCelestialLayer(scene)"]
+    B --> C["创建天球球壳"]
+    B --> D["创建太阳 sprite + 主定向光"]
+    B --> E["创建月亮 sprite"]
+
+    F["animate()"] --> G["updateCelestialLayer(now, camera, earth)"]
+    G --> H["Astronomy Engine 计算 Sun/Moon 方向"]
+    H --> I["更新 sun sprite / moon sprite 位置"]
+    H --> J["更新太阳 DirectionalLight 方向"]
+    J --> K["地球昼夜方向变化"]
+```
+
+## 风险与注意事项
+
+### 1. 星图投影方向容易反
+
+这会表现为：
+
+- 星图左右镜像
+- 赤经方向颠倒
+- 日月位置和背景对不上
+
+建议：
+
+- 先做一个开发调试模式
+- 显示赤经/赤纬参考点，快速校正纹理朝向
+
+### 2. 不要让天球跟随 Earth 旋转
+
+天球背景和日月必须属于 scene/world，而不是 `earthObj`。
+
+### 3. 不要每帧做重型天文计算
+
+真实位置更新应节流，否则会浪费 CPU。
+
+### 4. 月亮先求“方向正确”，再求“月相精致”
+
+月相属于第二步优化，不应阻塞 V1 上线。
+
+## 推荐实施顺序
+
+1. 新建 `celestial.js`
+2. 用星图球壳替换随机星点
+3. 接入 Astronomy Engine
+4. 加太阳/月亮 sprite
+5. 用太阳方向驱动主光
+6. 再决定要不要做月相和更严格参考系
+
+## 最终建议
+
+对于当前 Planet Earth，最稳妥的方案是：
+
+- 先做真实天球背景
+- 再做真实太阳/月亮方向
+- 再让太阳驱动地球受光
+- 暂时不做 Earth 参考系重构
+
+这样可以在不破坏现有 Earth 交互和图层系统的前提下，显著提升空间感、真实感和演示说服力。

docs/plans/earth-compute-center-bgp-style-plan.md

@@ -0,0 +1,372 @@
+# Earth Compute Center BGP-Style Plan
+
+## Goal
+
+这份文档定义如何按照 BGP 模块的产品方式，把“算力中心”提升为 Earth 上的一级能力。
+
+这里的“按 BGP 方式”指的是：
+
+- 有独立的数据语义和接口入口
+- 有独立的 Earth 图层与图例
+- 有独立的 hover / click / 选中态 / 详情卡
+- 有独立的统计口径与后续专题页扩展空间
+
+这里的“按 BGP 方式”不指：
+
+- 机械复制 BGP 的 anomaly / incident / collector 三层事件模型
+- 为静态算力设施强行引入不必要的复杂告警语义
+
+算力中心本质上更接近“长期基础设施分布层”，不是“高频动态异常层”。
+因此应该复用 BGP 的模块化方法，而不是照搬 BGP 的事件结构。
+
+## Why
+
+当前仓库里已经有算力相关基础：
+
+- 后端已有 `top500` 和 `epoch_ai_gpu` 数据采集
+- 可视化接口已有 `/api/v1/visualization/geo/supercomputers` 和 `/api/v1/visualization/geo/gpu-clusters`
+- Earth 信息卡已对 `supercomputer` 和 `gpu_cluster` 做了基础类型兼容
+
+但当前能力还停留在“数据可取到”的阶段，没有形成像 BGP 那样完整的可视化模块：
+
+- Earth 缺少独立的算力图层加载模块
+- 缺少算力 marker 体系和视觉层级
+- 缺少算力图例、统计、开关和搜索接入
+- 缺少与海缆、BGP、卫星的关系表达
+- 缺少算力专题页和后续告警/研判扩展入口
+
+所以当前真正的缺口不是“有没有数据”，而是“有没有产品级模块”。
+
+## Core Principle
+
+算力中心应当采用和 BGP 一致的模块化分层：
+
+1. 数据层：稳定的数据契约和 GeoJSON 输出
+2. 渲染层：独立的 Earth 图层、marker 和视觉状态管理
+3. 交互层：hover、click、锁定态、详情卡、图例和统计
+4. 扩展层：后续专题页、关系分析、告警和 AI 研判
+
+但语义上必须保持算力中心自身的特点：
+
+- `site / center` 是主对象，不是事件
+- `capacity / rank / vendor / operator / status` 是主信息，不是异常严重度
+- `distribution / concentration / dependency` 是后续分析方向，不是第一阶段必须项
+
+## Recommended Scope
+
+第一版“算力中心”建议统一承载两类对象：
+
+- `supercomputer`
+- `gpu_cluster`
+
+并在 Earth 上收口为一个主题层：`compute_centers`
+
+这样做有几个好处：
+
+- 用户看到的是统一的“算力基础设施”语义，而不是零散数据源
+- 后端仍可保留 `top500` 和 `epoch_ai_gpu` 的来源差异
+- 前端可以在一个图层里再细分两种 marker 语言
+
+## Current Gap
+
+和 BGP 对比，当前差距主要在下面几层。
+
+### 1. Data Contract Gap
+
+现在的算力 GeoJSON 还是通用 `collected_data` 输出思路，字段较轻：
+
+- `gpu_cluster` 只有基础名称和地点
+- `supercomputer` 只暴露一部分性能字段
+- 缺少统一的 `site_type / operator / capacity_band / source / updated_at / confidence`
+- 缺少统一的算力层聚合出口
+
+### 2. Earth Rendering Gap
+
+当前 Earth 里没有类似 `bgp.js` 的算力模块：
+
+- `constants.js` 没有算力 API 路径和视觉配置
+- `main.js` 没有算力加载、拾取、状态同步和 HUD 更新
+- `controls.js` 没有算力图层开关和启动加载优先级
+- `layer-startup-tasks.js` 没有算力启动任务
+- `legend.js` / `ui.js` 没有算力统计与图例模式
+
+### 3. Interaction Gap
+
+虽然 `info-card.js` 支持基础字段，但还没有形成 BGP 那种完整交互链路：
+
+- 没有 hover / selected / dimmed 的视觉状态
+- 没有算力对象专属 tooltip 与摘要文案
+- 没有锁定后与其他基础设施的联动高亮
+- 没有搜索、统计卡和详情组织方式
+
+### 4. Product Expansion Gap
+
+当前还没有“算力中心”专题页与分析语义：
+
+- 没有全球分布/国家聚合/厂商聚合视图
+- 没有算力与海缆/BGP/区域的关系表达
+- 没有 AI brief / assessment 的后续落点
+
+## Architecture Direction
+
+推荐把算力中心做成“BGP 同级能力”，但采用更适合静态基础设施的结构。
+
+### Backend
+
+建议新增统一聚合接口，例如：
+
+- `/api/v1/visualization/geo/compute-centers`
+
+它的职责是把：
+
+- `top500`
+- `epoch_ai_gpu`
+
+统一转换成一个主题层输出，同时保留对象细分类型：
+
+- `site_type: supercomputer | gpu_cluster`
+
+建议统一字段至少包括：
+
+- `id`
+- `name`
+- `site_type`
+- `country`
+- `city`
+- `latitude`
+- `longitude`
+- `operator`
+- `vendor`
+- `capacity_value`
+- `capacity_unit`
+- `capacity_band`
+- `rank`
+- `source`
+- `updated_at`
+- `location_precision`
+- `geography_mode`
+- `is_estimated`
+- `estimated_reason`
+- `metadata`
+
+这里建议优先做“统一聚合出口”，而不是一开始就新增独立数据库表。
+
+原因：
+
+- 当前源数据更新频率低，先复用 `collected_data` 成本更低
+- 可以先把 Earth 产品体验做完整
+- 如果后续要做历史趋势、关系推断、告警，再评估是否拆成独立模型
+
+### Frontend Earth
+
+建议新增独立模块，例如：
+
+- `frontend/public/earth/js/compute-centers.js`
+
+职责参照 `bgp.js`：
+
+- 拉取算力中心 GeoJSON
+- 创建 marker
+- 管理 hover / selected / dimmed 状态
+- 输出图例项
+- 输出统计摘要
+- 提供 overlay 和详情格式化辅助函数
+
+推荐视觉分层：
+
+1. `supercomputer` 用更稳定、更规整的设施型符号
+2. `gpu_cluster` 用更活跃、更现代的密度型符号
+3. 选中态通过 halo / ring / related infrastructure highlight 表达
+
+视觉上应避免把算力中心做成“BGP 事件点”那种高频脉冲风格。
+它应该更像长期存在的高价值设施。
+
+## Phases
+
+## Phase 1: Unified Earth Layer
+
+目标：
+
+- 先把算力中心做成 Earth 上可用、可点、可解释的一级图层
+
+工作项：
+
+- 新增统一算力 GeoJSON 接口
+- 新增 `compute-centers.js`
+- 在 `constants.js` 增加 API 路径和视觉配置
+- 在 `controls.js` 增加算力图层开关与启动元数据
+- 在 `layer-startup-tasks.js` 增加算力启动加载任务
+- 在 `main.js` 接入算力拾取、hover、click、锁定态和 HUD 统计
+- 在 `ui.js` / `legend.js` / `index.html` 增加算力统计与图例入口
+- 在 `info-card.js` 提升算力详情字段组织
+- 对无法精确定位、但可按国家或弱线索推测的大概位置，仍然生成地图点位
+- 这类对象必须带显式“估算位置”状态，例如图标问号角标与详情说明
+
+完成标准：
+
+- Earth 上能独立显示/隐藏算力中心
+- 两类对象有可区分的视觉表达
+- hover / click / 详情卡 / 图例 / 统计全部打通
+- 精确位置与估算位置在图标或文案上可区分，不会误导为同一精度
+- 不干扰现有海缆、卫星、BGP 的交互链路
+
+## Phase 2: Relationship Layer
+
+目标：
+
+- 让算力中心不只是“点”，而是和其他基础设施产生上下文关系
+
+工作项：
+
+- 建立算力中心与国家/区域聚合摘要
+- 增加与附近海缆登陆点的关系提示
+- 增加与 BGP 事件/观测范围的空间邻近提示
+- 增加与卫星覆盖或区域连通性的实验性提示
+
+完成标准：
+
+- 点击算力中心时，用户能看到“它和哪些基础设施相关”
+- 信息表达以辅助判断为主，不做夸张推断
+
+## Phase 3: Compute Center Observatory
+
+目标：
+
+- 把算力中心从 Earth 图层扩展成独立专题观测能力
+
+工作项：
+
+- 新增算力中心专题页
+- 提供国家/厂商/类型/容量分布统计
+- 支持列表、筛选、详情和历史快照
+- 预留 AI brief / assessment 入口
+
+完成标准：
+
+- 算力中心不再只是 Earth 上的视觉点位
+- 能作为独立业务上下文进入日常观察与研判
+
+## Phase 4: Alerts And Assessment
+
+目标：
+
+- 在不滥造“假动态告警”的前提下，引入真正有价值的变化感知
+
+候选方向：
+
+- 新增大规模算力中心
+- 既有中心容量显著变化
+- 国家/区域集中度显著变化
+- 高价值中心与关键网络基础设施关系变化
+
+完成标准：
+
+- 告警来自可解释的结构变化
+- 不把静态数据硬做成噪声式实时事件流
+
+## Implementation Notes
+
+建议按下面顺序推进：
+
+1. 先统一 GeoJSON 契约
+2. 再做 Earth 独立模块和图层开关
+3. 再补详情卡、图例和统计
+4. 最后才做关系层和专题页
+
+这样可以避免一开始把范围摊得过大。
+
+## Unknown Location Strategy
+
+由于部分算力数据源不会直接提供经纬度，未知位置补全不能只依赖“继续找 API 字段”。
+更稳妥的方式是做成一条分层富化链路，而不是单一猜测规则。
+
+推荐按下面优先级推进：
+
+1. 直接源信息
+
+- 源记录显式给出 `latitude / longitude`
+- 源记录给出 `city / region / facility / campus / operator`
+- 源页面详情、内嵌 JSON、结构化元数据、新闻稿链接里能抽出地点线索
+
+2. 名称与机构归一化
+
+- 建立 `canonical_name / aliases / operator / facility` 归一化表
+- 把 `cluster name`、`operator`、`campus name` 归一到同一个实体
+- 优先解决同一对象多写法导致的命中失败，而不是先扩大猜测范围
+
+3. 本地位置注册表
+
+- 用仓库内可维护的 registry 保存高价值对象的位置知识
+- 每条记录至少包含：`canonical_name`、`aliases`、`operator`、`country`、`region`、`city`、`lat`、`lon`、`confidence`、`source_note`
+- 转换层优先读取 registry，避免地点知识长期散落在转换代码里
+
+4. 分层回退定位
+
+- `precise`
+- `estimated_site`
+- `estimated_city`
+- `estimated_region`
+- `estimated_national_hub`
+- `estimated_country`
+
+这里建议把“国家内主要算力城市”作为国家质心之前的一层。
+例如没有美国精确位置时，优先考虑已知的主要算力/数据中心城市候选，而不是直接落在几何质心。
+
+5. 候选证据富化
+
+- 如果源 API 无地点信息，可以允许采集链路读取公开辅助证据
+- 例如机构官网、数据中心介绍页、新闻稿、百科型页面、公开 PDF
+- 但只提取“地点线索”，不把外部页面上的经纬度当真值直接写回
+
+6. 人工校验闭环
+
+- 对高价值且仍然未知的对象输出待核验清单
+- 把人工确认结果回写到位置注册表
+- 后续采集继续优先复用这层人工确认结果
+
+### Additional Solution Paths
+
+除了静态映射表，还可以考虑下面这些办法：
+
+- 基于国家和运营方建立“主要园区候选集”，用稳定散列把同国未知节点分散到若干可信城市，而不是全部压到一个点
+- 基于数据中心/云厂商公开 region 列表建立 `operator -> city set` 候选映射，用于云 GPU 集群类对象
+- 把“估算依据”结构化，例如 `matched_alias`、`matched_operator`、`matched_city_text`、`fallback_country_hub`
+- 给位置补全增加 `last_verified_at`，便于后续按时间重新校验老旧映射
+- 单独维护“不可可靠定位”状态；这类对象仍可在国家级聚合统计中出现，但可以允许用户在地图上过滤掉
+- 后续如果你们愿意投入更多，可把这条链路做成小型 enrichment pipeline，而不是仅在 API 转换时临时判断
+
+## Non-Goals
+
+第一阶段不建议做这些内容：
+
+- 不复制 BGP 巡航模式到算力中心
+- 不先做复杂实时 websocket 推送
+- 不先引入独立 `compute_center_incident` 一类模型
+- 不先做全量 AI 分析面板
+
+原因是算力中心的第一需求是“被看清楚”，不是“被实时播报”。
+但“被看清楚”不等于“只显示精确坐标对象”。
+对于没有精确经纬度、但能推测到国家或区域级位置的算力中心，应优先以上图并标注估算状态的方式处理，而不是直接在地图上消失。
+
+## Acceptance Checklist
+
+- 后端存在统一的算力中心 GeoJSON 出口
+- Earth 有独立算力图层模块，而不是散落在 `main.js`
+- 页面上有清晰的算力开关、图例和统计
+- `supercomputer` 和 `gpu_cluster` 在视觉和详情上都可区分
+- 估算位置对象在地图和详情中都有明确状态提示
+- 现有 BGP / 海缆 / 卫星功能无回归
+- 代码结构上为后续专题页和关系分析留出了明确扩展点
+
+## Summary
+
+这项工作的本质不是“再多画几个点”。
+
+它应该把算力中心从已有数据源，升级成与 BGP 同级的 Earth 观测主题：
+
+- 有独立语义
+- 有独立图层
+- 有独立交互
+- 有后续分析扩展能力
+
+推荐先完成 Phase 1，把算力中心做成真正可用的 Earth 一级模块，再继续推进关系层和专题页。