docs: translate daily-news-report skill and update registry files

This commit is contained in:
Ianj332
2026-01-27 15:24:44 -06:00
parent 3446382851
commit 2a92eba60e
2 changed files with 173 additions and 174 deletions

View File

@@ -1,7 +1,7 @@
---
name: daily-news-report
description: 基于预设 URL 列表抓取内容,筛选高质量技术信息并生成每日 Markdown 报告。
argument-hint: [可选: 日期]
description: Scrapes content based on a preset URL list, filters high-quality technical information, and generates daily Markdown reports.
argument-hint: [optional: date]
disable-model-invocation: false
user-invocable: true
allowed-tools: Task, WebFetch, Read, Write, Bash(mkdir*), Bash(date*), Bash(ls*), mcp__chrome-devtools__*
@@ -9,93 +9,93 @@ allowed-tools: Task, WebFetch, Read, Write, Bash(mkdir*), Bash(date*), Bash(ls*)
# Daily News Report v3.0
> **架构升级**:主 Agent 调度 + SubAgent 执行 + 浏览器抓取 + 智能缓存
> **Architecture Upgrade**: Main Agent Orchestration + SubAgent Execution + Browser Scraping + Smart Caching
## 核心架构
## Core Architecture
```
┌─────────────────────────────────────────────────────────────────────┐
Agent (Orchestrator)
职责:调度、监控、评估、决策、汇总
Main Agent (Orchestrator) │
Role: Scheduling, Monitoring, Evaluation, Decision, Aggregation
├─────────────────────────────────────────────────────────────────────┤
│ │
│ ┌──────────┐ ┌──────────┐ ┌──────────┐ ┌──────────┐ │
│ │ 1. 初始化 │ → │ 2. 调度 │ → │ 3. 监控 │ → │ 4. 评估 │ │
│ │ 读取配置 │ │ 分发任务 │ │ 收集结果 │ │ 筛选排序 │ │
│ └──────────┘ └──────────┘ └──────────┘ └──────────┘ │
│ │ │ │ │
│ ▼ ▼ ▼ │
│ ┌──────────┐ ┌──────────┐ ┌──────────┐ ┌──────────┐ │
│ │ 5. 决策 │ ← │ 够20条 │ │ 6. 生成 │ → │ 7. 更新 │ │
│ │ 继续/停止 │ │ Y/N │ │ 日报文件 │ │ 缓存统计 │ │
│ └──────────┘ └──────────┘ └──────────┘ └──────────┘ │
│ ┌─────────────┐ ┌─────────────┐ ┌─────────────┐ ┌─────────────┐ │
│ │ 1. Init │ → │ 2. Dispatch │ → │ 3. Monitor │ → │ 4. Evaluate │ │
│ │ Read Config │ │ Assign Tasks│ │ Collect Res │ │ Filter/Sort │ │
│ └─────────────┘ └─────────────┘ └─────────────┘ └─────────────┘ │
│ │ │ │
│ ▼ ▼ │
│ ┌─────────────┐ ┌─────────────┐ ┌─────────────┐ ┌─────────────┐ │
│ │ 5. Decision │ ← │ Enough 20? │ │ 6. Generate │ → │ 7. Update │ │
│ │ Cont/Stop │ │ Y/N │ │ Report File │ │ Cache Stats │ │
│ └─────────────┘ └─────────────┘ └─────────────┘ └─────────────┘ │
│ │
└──────────────────────────────────────────────────────────────────────┘
调度 返回结果
DispatchReturn Results
┌─────────────────────────────────────────────────────────────────────┐
│ SubAgent 执行层
│ SubAgent Execution Layer
├─────────────────────────────────────────────────────────────────────┤
│ │
│ ┌─────────────┐ ┌─────────────┐ ┌─────────────┐ │
│ │ Worker A │ │ Worker B │ │ Browser │ │
│ │ (WebFetch) │ │ (WebFetch) │ │ (Headless) │ │
│ │ Tier1 Batch │ │ Tier2 Batch │ │ JS渲染页面 │ │
│ │ Tier1 Batch │ │ Tier2 Batch │ │ JS Render │ │
│ └─────────────┘ └─────────────┘ └─────────────┘ │
│ ↓ ↓ ↓ │
│ ┌─────────────────────────────────────────────────────────────┐ │
│ │ 结构化结果返回 │ │
│ │ Structured Result Return │ │
│ │ { status, data: [...], errors: [...], metadata: {...} } │ │
│ └─────────────────────────────────────────────────────────────┘ │
│ │
└─────────────────────────────────────────────────────────────────────┘
```
## 配置文件
## Configuration Files
本 Skill 使用以下配置文件:
This skill uses the following configuration files:
| 文件 | 用途 |
|------|------|
| `sources.json` | 信息源配置、优先级、抓取方法 |
| `cache.json` | 缓存数据、历史统计、去重指纹 |
| File | Purpose |
|------|---------|
| `sources.json` | Source configuration, priorities, scrape methods |
| `cache.json` | Cached data, historical stats, deduplication fingerprints |
## 执行流程详解
## Execution Process Details
### Phase 1: 初始化
### Phase 1: Initialization
```yaml
步骤:
1. 确定日期(用户参数或当前日期)
2. 读取 sources.json 获取源配置
3. 读取 cache.json 获取历史数据
4. 创建输出目录 NewsReport/
5. 检查今日是否已有部分报告(追加模式)
Steps:
1. Determine date (user argument or current date)
2. Read sources.json for source configurations
3. Read cache.json for historical data
4. Create output directory NewsReport/
5. Check if a partial report exists for today (append mode)
```
### Phase 2: 调度 SubAgent
### Phase 2: Dispatch SubAgents
**策略**:并行调度,分批执行,早停机制
**Strategy**: Parallel dispatch, batch execution, early stopping mechanism
```yaml
第1波 (并行):
Wave 1 (Parallel):
- Worker A: Tier1 Batch A (HN, HuggingFace Papers)
- Worker B: Tier1 Batch B (OneUsefulThing, Paul Graham)
等待结果 → 评估数量
Wait for results → Evaluate count
如果 < 15 条高质量:
第2波 (并行):
If < 15 high-quality items:
Wave 2 (Parallel):
- Worker C: Tier2 Batch A (James Clear, FS Blog)
- Worker D: Tier2 Batch B (HackerNoon, Scott Young)
如果仍 < 20 :
第3波 (浏览器):
- Browser Worker: ProductHunt, Latent Space (需要JS渲染)
If still < 20 items:
Wave 3 (Browser):
- Browser Worker: ProductHunt, Latent Space (Require JS rendering)
```
### Phase 3: SubAgent 任务格式
### Phase 3: SubAgent Task Format
每个 SubAgent 接收的任务格式:
Task format received by each SubAgent:
```yaml
task: fetch_and_extract
@@ -109,127 +109,127 @@ sources:
output_schema:
items:
- source_id: string # 来源标识
title: string # 标题
summary: string # 2-4句摘要
key_points: string[] # 最多3个要点
url: string # 原文链接
keywords: string[] # 关键词
quality_score: 1-5 # 质量评分
- source_id: string # Source Identifier
title: string # Title
summary: string # 2-4 sentence summary
key_points: string[] # Max 3 key points
url: string # Original URL
keywords: string[] # Keywords
quality_score: 1-5 # Quality Score
constraints:
filter: "前沿技术/高深技术/提效技术/实用资讯"
exclude: "泛科普/营销软文/过度学术化/招聘帖"
filter: "Cutting-edge Tech/Deep Tech/Productivity/Practical Info"
exclude: "General Science/Marketing Puff/Overly Academic/Job Posts"
max_items_per_source: 10
skip_on_error: true
return_format: JSON
```
### Phase 4: Agent 监控与反馈
### Phase 4: Main Agent Monitoring & Feedback
Agent 职责:
Main Agent Responsibilities:
```yaml
监控:
- 检查 SubAgent 返回状态 (success/partial/failed)
- 统计收集到的条目数量
- 记录每个源的成功率
Monitoring:
- Check SubAgent return status (success/partial/failed)
- Count collected items
- Record success rate per source
反馈循环:
- 如果某 SubAgent 失败,决定是否重试或跳过
- 如果某源持续失败,标记为禁用
- 动态调整后续批次的源选择
Feedback Loop:
- If a SubAgent fails, decide whether to retry or skip
- If a source fails persistently, mark as disabled
- Dynamically adjust source selection for subsequent batches
决策:
- 条目数 >= 25 且高质量 >= 20 → 停止抓取
- 条目数 < 15 → 继续下一批
- 所有批次完成但 < 20 → 用现有内容生成(宁缺毋滥)
Decision:
- Items >= 25 AND HighQuality >= 20 → Stop scraping
- Items < 15 → Continue to next batch
- All batches done but < 20 → Generate with available content (Quality over Quantity)
```
### Phase 5: 评估与筛选
### Phase 5: Evaluation & Filtering
```yaml
去重:
- 基于 URL 完全匹配
- 基于标题相似度 (>80% 视为重复)
- 检查 cache.json 避免与历史重复
Deduplication:
- Exact URL match
- Title similarity (>80% considered duplicate)
- Check cache.json to avoid history duplicates
评分校准:
- 统一各 SubAgent 的评分标准
- 根据来源可信度调整权重
- 手动标注的高质量源加分
Score Calibration:
- Unify scoring standards across SubAgents
- Adjust weights based on source credibility
- Bonus points for manually curated high-quality sources
排序:
- quality_score 降序
- 同分按来源优先级排序
- 截取 Top 20
Sorting:
- Descending order by quality_score
- Sort by source priority if scores are equal
- Take Top 20
```
### Phase 6: 浏览器抓取 (MCP Chrome DevTools)
### Phase 6: Browser Scraping (MCP Chrome DevTools)
对于需要 JS 渲染的页面,使用无头浏览器:
For pages requiring JS rendering, use a headless browser:
```yaml
流程:
1. 调用 mcp__chrome-devtools__new_page 打开页面
2. 调用 mcp__chrome-devtools__wait_for 等待内容加载
3. 调用 mcp__chrome-devtools__take_snapshot 获取页面结构
4. 解析 snapshot 提取所需内容
5. 调用 mcp__chrome-devtools__close_page 关闭页面
Process:
1. Call mcp__chrome-devtools__new_page to open page
2. Call mcp__chrome-devtools__wait_for to wait for content load
3. Call mcp__chrome-devtools__take_snapshot to get page structure
4. Parse snapshot to extract required content
5. Call mcp__chrome-devtools__close_page to close page
适用场景:
Applicable Scenarios:
- ProductHunt (403 on WebFetch)
- Latent Space (Substack JS 渲染)
- 其他 SPA 应用
- Latent Space (Substack JS rendering)
- Other SPA applications
```
### Phase 7: 生成日报
### Phase 7: Generate Report
```yaml
输出:
- 目录: NewsReport/
- 文件名: YYYY-MM-DD-news-report.md
- 格式: 标准 Markdown
Output:
- Directory: NewsReport/
- Filename: YYYY-MM-DD-news-report.md
- Format: Standard Markdown
内容结构:
- 标题 + 日期
- 统计摘要(源数量、收录数量)
- 20条高质量内容(按模板)
- 生成信息(版本、时间戳)
Content Structure:
- Title + Date
- Statistical Summary (Source count, items collected)
- 20 High-Quality Items (Template based)
- Generation Info (Version, Timestamps)
```
### Phase 8: 更新缓存
### Phase 8: Update Cache
```yaml
更新 cache.json:
- last_run: 记录本次运行信息
- source_stats: 更新各源统计数据
- url_cache: 添加已处理的 URL
- content_hashes: 添加内容指纹
- article_history: 记录收录文章
Update cache.json:
- last_run: Record this run info
- source_stats: Update stats per source
- url_cache: Add processed URLs
- content_hashes: Add content fingerprints
- article_history: Record included articles
```
## SubAgent 调用示例
## SubAgent Call Examples
### 使用 general-purpose Agent
### Using general-purpose Agent
由于自定义 agent 需要 session 重启才能发现,可以使用 general-purpose 并注入 worker prompt
Since custom agents require session restart to be discovered, use general-purpose and inject worker prompts:
```
Task 调用:
Task Call:
subagent_type: general-purpose
model: haiku
prompt: |
你是一个无状态的执行单元。只做被分配的任务,返回结构化 JSON
You are a stateless execution unit. Only do the assigned task and return structured JSON.
任务:抓取以下 URL 并提取内容
Task: Scrape the following URLs and extract content
URLs:
- https://news.ycombinator.com (提取 Top 10)
- https://huggingface.co/papers (提取高投票论文)
- https://news.ycombinator.com (Extract Top 10)
- https://huggingface.co/papers (Extract top voted papers)
输出格式:
Output Format:
{
"status": "success" | "partial" | "failed",
"data": [
@@ -247,17 +247,17 @@ Task 调用:
"metadata": { "processed": 2, "failed": 0 }
}
筛选标准:
- 保留:前沿技术/高深技术/提效技术/实用资讯
- 排除:泛科普/营销软文/过度学术化/招聘帖
Filter Criteria:
- Keep: Cutting-edge Tech/Deep Tech/Productivity/Practical Info
- Exclude: General Science/Marketing Puff/Overly Academic/Job Posts
直接返回 JSON不要解释。
Return JSON directly, no explanation.
```
### 使用 worker Agent(需重启 session
### Using worker Agent (Requires session restart)
```
Task 调用:
Task Call:
subagent_type: worker
prompt: |
task: fetch_and_extract
@@ -274,37 +274,36 @@ Task 调用:
- keywords: string[]
- quality_score: 1-5
constraints:
filter: 前沿技术/高深技术/提效技术/实用资讯
exclude: 泛科普/营销软文/过度学术化
filter: Cutting-edge Tech/Deep Tech/Productivity/Practical Info
exclude: General Science/Marketing Puff/Overly Academic
```
## 输出模板
## Output Template
```markdown
# Daily News ReportYYYY-MM-DD
# Daily News Report (YYYY-MM-DD)
> 本日筛选自 N 个信息源,共收录 20 条高质量内容
> 生成耗时: X 分钟 | 版本: v3.0
> Curated from N sources today, containing 20 high-quality items
> Generation Time: X min | Version: v3.0
>
> **Warning**: Sub-agent 'worker' not detected. Running in generic mode (Serial Execution). Performance might be degraded.
> **警告**:未检测到 Sub-agent 'worker'。正在以通用模式(串行执行)运行。性能可能会受影响。
---
## 1. 标题
## 1. Title
- **摘要**2-4 行概述
- **要点**
1. 要点一
2. 要点二
3. 要点三
- **来源**[链接](URL)
- **关键词**`keyword1` `keyword2` `keyword3`
- **评分**⭐⭐⭐⭐⭐ (5/5)
- **Summary**: 2-4 lines overview
- **Key Points**:
1. Point one
2. Point two
3. Point three
- **Source**: [Link](URL)
- **Keywords**: `keyword1` `keyword2` `keyword3`
- **Score**: ⭐⭐⭐⭐⭐ (5/5)
---
## 2. 标题
## 2. Title
...
---
@@ -313,45 +312,45 @@ Task 调用:
*Sources: HN, HuggingFace, OneUsefulThing, ...*
```
## 约束与原则
## Constraints & Principles
1. **宁缺毋滥**:低质量内容不进入日报
2. **早停机制**:够 20 条高质量就停止抓取
3. **并行优先**:同一批次的 SubAgent 并行执行
4. **失败容错**:单个源失败不影响整体流程
5. **缓存复用**:避免重复抓取相同内容
6. ** Agent 控制**:所有决策由主 Agent 做出
7. **Fallback Awareness**:检测 sub-agent 可用性,不可用时优雅降级
1. **Quality over Quantity**: Low-quality content does not enter the report.
2. **Early Stop**: Stop scraping once 20 high-quality items are reached.
3. **Parallel First**: SubAgents in the same batch execute in parallel.
4. **Fault Tolerance**: Failure of a single source does not affect the whole process.
5. **Cache Reuse**: Avoid re-scraping the same content.
6. **Main Agent Control**: All decisions are made by the Main Agent.
7. **Fallback Awareness**: Detect sub-agent availability, gracefully degrade if unavailable.
## 预期性能
## Expected Performance
| 场景 | 预期时间 | 说明 |
|------|----------|------|
| 最优情况 | ~2 分钟 | Tier1 足够,无需浏览器 |
| 正常情况 | ~3-4 分钟 | 需要 Tier2 补充 |
| 需要浏览器 | ~5-6 分钟 | 包含 JS 渲染页面 |
| Scenario | Expected Time | Note |
|---|---|---|
| Optimal | ~2 mins | Tier1 sufficient, no browser needed |
| Normal | ~3-4 mins | Requires Tier2 supplement |
| Browser Needed | ~5-6 mins | Includes JS rendered pages |
## 错误处理
## Error Handling
| 错误类型 | 处理方式 |
|----------|----------|
| SubAgent 超时 | 记录错误,继续下一个 |
| 403/404 | 标记禁用,更新 sources.json |
| 内容提取失败 | 返回原始内容,主 Agent 决定 |
| 浏览器崩溃 | 跳过该源,记录日志 |
| Error Type | Handling |
|---|---|
| SubAgent Timeout | Log error, continue to next |
| Source 403/404 | Mark disabled, update sources.json |
| Extraction Failed | Return raw content, Main Agent decides |
| Browser Crash | Skip source, log entry |
## 兼容性与兜底 (Compatibility & Fallback)
## Compatibility & Fallback
为了确保在不同 Agent 环境下的可用性,必须执行以下检查:
To ensure usability across different Agent environments, the following checks must be performed:
1. **环境检查**:
- Phase 1 初始化阶段,尝试检测 `worker` sub-agent 是否存在。
- 如果不存在(或未安装相关插件),自动切换到 **串行执行模式 (Serial Mode)**
1. **Environment Check**:
- In Phase 1 initialization, attempt to detect if `worker` sub-agent exists.
- If not exists (or plugin not installed), automatically switch to **Serial Execution Mode**.
2. **串行执行模式**:
- 不使用 parallel block
- Agent 依次执行每个源的抓取任务。
- 虽然速度较慢,但保证基本功能可用。
2. **Serial Execution Mode**:
- Do not use parallel block.
- Main Agent executes scraping tasks for each source sequentially.
- Slower, but guarantees basic functionality.
3. **用户提示**:
- 必须在生成的日报开头(引用块部分)包含明显的警告信息,提示用户当前正在运行于降级模式。
3. **User Alert**:
- MUST include a clear warning in the generated report header indicating the current degraded mode.