功能描述

内容捕手 (Content Hunter) - 短视频平台热门内容抓取机器人。支持小红书、抖音、B站，可分批抓取热门内容并自动生成汇报。/Content Hunter - Hot content crawler for Xiaohongshu, Douyin, Bilibili. Supports batch s...

使用说明 (SKILL.md)

内容捕手 / Content Hunter

Name: 内容捕手 Content Hunter
Author: judefluen-coder

概述 / Overview

持续刷小红书、抖音，B站3个平台，每次每平台抓取40条热门内容并保存，晚上8点生成汇报。/Scrapes 3 platforms (Xiaohongshu, Douyin, Bilibili), 40 items each, generates daily report at 8 PM.

一、浏览器标签管理 / Browser Tab Management

平台 / Platform	网站URL / URL	提取要素 / Data Fields	备注 / Notes
小红书 / Xiaohongshu	https://www.xiaohongshu.com	标题、作者、点赞、收藏、评论、话题、内容总结 / Title, Author, Likes, Saves, Comments, Tags, Summary	需要登录 / Login required
抖音 / Douyin	https://www.douyin.com	标题、作者、点赞、话题、内容总结 / Title, Author, Likes, Tags, Summary
B站 / Bilibili	https://www.bilibili.com	标题、UP主、播放、弹幕、点赞、投币、收藏、字幕、内容总结 / Title, UP, Views, Danmaku, Likes, Coins, Favs, Subtitles, Summary	尝试提取字幕 / Try extract subtitles

每次每个平台抓取40条热门内容 / Scrape 40 hot items per platform each time

二、存储位置 / Storage Location

工作空间 / Workspace：~/.openclaw/workspace/content-hunter/

重要：每次任务新建独立子文件夹，不覆盖历史数据！ / Important: Create new subfolder for each task, NEVER overwrite historical data!

content-hunter/
├── task-2026-03-11-0900/     # 任务1：按时间命名
│   ├── xiaohongshu.md
│   ├── douyin.md
│   ├── bilibili.md
│   └── summary.md              # 本次任务汇报
├── task-2026-03-11-0930/     # 任务2
│   ├── xiaohongshu.md
│   ├── douyin.md
│   ├── bilibili.md
│   └── summary.md
└── task-2026-03-11-2000/     # 晚8点汇报任务
    └── summary.md

文件夹命名规则：task-YYYY-MM-DD-HHMM

每次任务执行前：

创建新文件夹：task-$(date +%Y-%m-%d-%H%M)
在新文件夹中抓取和保存数据

三、记录格式 / Data Format

重要：每条记录必须包含以下所有字段，缺一不可！ / Important: Every record MUST have ALL these fields!

小红书 / Xiaohongshu

### 第X条 / Item #X
- 标题 / Title: xxx
- 作者 / Author: xxx
- 点赞 / Likes: xxx
- 收藏 / Saves: xxx
- 评论 / Comments: xxx
- 话题 / Tags: #xxx #xxx
- 内容总结 / Summary: 详细描述... (3-5句话 / 3-5 sentences)

抖音 / Douyin

### 第X条 / Item #X
- 标题 / Title: xxx
- 作者 / Author: @xxx
- 点赞 / Likes: xxx
- 话题 / Tags: #xxx
- 内容总结 / Summary: 详细描述... (3-5句话 / 3-5 sentences)

B站 / Bilibili

### 第X条 / Item #X
- 标题 / Title: xxx
- UP主 / UP: xxx
- 播放 / Views: xxx
- 弹幕 / Danmaku: xxx
- 点赞 / Likes: xxx
- 投币 / Coins: xxx
- 收藏 / Favs: xxx
- 字幕 / Subtitles: 有/无 / Yes/No
- 内容总结 / Summary: 详细描述... (3-5句话 / 3-5 sentences)

四、执行流程 / Execution Flow

4.1 抓取要求 / Scraping Requirement

每次每个平台抓取40条热门内容 / Scrape 40 hot items per platform each time

4.2 初始化 / Initialization

创建新任务文件夹 / Create new task folder：mkdir -p ~/.openclaw/workspace/content-hunter/task-$(date +%Y-%m-%d-%H%M)/
用 browser.open 打开3个平台 / Open 3 platforms via browser.open

4.3 循环执行 / Loop Execution

browser.snapshot() 获取页面内容 / Get page content
必须提取所有字段！每条必须包含：标题、作者、热度、内容总结 / Must extract ALL fields: title, author, views/likes, content summary
保存到当前任务文件夹 / Save to current task folder：~/.openclaw/workspace/content-hunter/task-YYYY-MM-DD-HHMM/
等待3-5分钟 / Wait 3-5 minutes
⚠️ 抓取过程中不发送任何进度消息到群聊，只保存数据 / ⚠️ Do NOT send any progress messages to group chats during scraping, only save data
汇报任务执行完毕后，停止抓取，等待下次定时任务 / After report is sent, stop scraping and wait for next scheduled task

4.4 汇报数据范围 / Report Data Scope

重要：汇报时需要汇总当日所有抓取任务的数据，而非仅最近一次！ / Important: Report must summarize ALL data from all tasks on that day, not just the latest one!

汇报生成流程：

读取当日所有task文件夹 / Read all task folders for the day
合并各平台的抓取数据 / Merge data from all platforms
去重后生成汇报 / Deduplicate and generate report
汇总数据量 = 所有任务的累加总和 / Total = sum of all tasks

4.5 B站字幕提取 / Bilibili Subtitle Extraction

使用 bilibili-youtube-watcher 技能的 get_transcript.py 脚本提取字幕

五、晚上汇报 / Daily Report

汇报流程 / Report Flow

读取数据：从最新的task文件夹读取各平台md文件 / Read from latest task folder
生成汇报 / Generate report
发送汇报到群里 / Send report to group
汇报后自动清理：执行 openclaw cron list 找到所有内容捕手相关 cron（名称含"内容捕手"或"hunter"），逐一删除
汇报后停止，等待下次任务 / After cleanup, stop and wait for next task

汇报内容 / Report Contents

发原始数据文件链接 / Send raw data links
热门内容TOP20（必须包含内容总结！）/ TOP 20 (MUST include content summary!)
- 格式：按平台分类，每条包含：标题 | 热度 | 内容总结（1-2句话）
- 例如：小红书暖心正能量：标题 → 热度 | 内容总结
- Format: By platform, each item: Title | Views/Likes | Summary
各平台趋势 + 整体趋势 / Platform trends + Overall trends
推荐深度阅读 / Recommended deep reads
呱呱感悟 / Guagua's thoughts
关闭浏览器 / Close browser tabs

六、可用平台 / Supported Platforms

平台 / Platform	状态 / Status
小红书 / Xiaohongshu	✅
抖音 / Douyin	✅
B站 / Bilibili	✅

七、使用方式 / Usage

7.1 触发方式 / How to Trigger

方式一：手动对话 / Method 1: Manual Chat

"帮我刷短视频平台" / "Scrape short video platforms"
"抓取热门内容" / "Get trending content"
"生成内容汇报" / "Generate content report"
"启用内容捕手" / "Enable content hunter"

方式二：定时任务 / Method 2: Cron Job (Recommended)

定时任务需要手动创建，参考以下配置：

抓取任务：每30分钟执行一次（9:00-20:00），静默执行不发群消息
汇报任务：晚上20:00执行，发送到小红书群

⚠️ 绝对原则：内容捕手的所有任务和汇报，只允许发送到小红书呱呱群（oc_d21f6b6f9bd843444622c8e221134f47），禁止发送到任何其他群！

7.2 可配置参数 / Configurable Parameters

参数 / Parameter	说明 / Description	默认值 / Default	设置方式 / How to Set
平台 / Platforms	要抓取的平台 / Platforms to scrape	全部3个 / All 3	"只刷抖音" / "Only Douyin"
数量 / Count	每平台抓取条数 / Items per platform	40条 / 40	"每个刷30条" / "30 items each"
汇报时间 / Report Time	生成汇报的时间 / When to generate report	20:00 / 8 PM	"晚上6点汇报" / "Report at 6 PM"
存储路径	数据保存目录	~/workspace/content-hunter/task-YYYY-MM-DD-HHMM/	一般不改

7.3 实际存储示例 / Storage Example

任务执行后，文件夹里有什么？

例如：用户说"帮我刷短视频平台"

~/.openclaw/workspace/content-hunter/
└── task-2026-03-11-2100/           # 任务文件夹（按时间命名）
    ├── xiaohongshu.md              # 小红书数据（40条，每条含标题、作者、点赞、内容总结）
    ├── douyin.md                    # 抖音数据
    ├── bilibili.md                 # B站数据
    └── summary.md                  # 汇报文件

每个md文件内容示例：

# 小红书热门内容

### 第1条
- 标题: 最近大家都在安装openclaw
- 作者: 美丽的一天
- 点赞: 1000
- 内容总结: 介绍OpenClaw的安装和使用方法...

### 第2条
- 标题: ...
...
（共40条 / 40 items total）

7.4 汇报内容示例 / Report Example

晚上8点汇报长这样：

📊 2026年3月11日 内容捕手汇报

🔥 热门内容TOP20
【小红书】
1. 标题 → 热度 | 内容总结
2. ...

【B站】
1. 标题 → 热度 | 内容总结
...

📈 各平台趋势总结
- 小红书：xxx
- 抖音：xxx
...

🎯 整体趋势
1. xxx
2. xxx

📖 推荐深度阅读
- xxx

🐸 呱呱感悟
xxx

7.5 对话示例 / Chat Examples

用户A：帮我刷短视频平台
→ AI：好的，启动内容捕手，默认抓取3个平台，每平台20条
/User A: Scrape video platforms
→ AI: OK, starting Content Hunter, 3 platforms, 20 items each

用户B：我只刷抖音和B站，每个刷30条
→ AI：好的，只抓取抖音和B站，每平台30条
/User B: Only Douyin and Bilibili, 30 each
→ AI: OK, Douyin + Bilibili, 30 items each

用户C：每天早上9点开始抓取，晚上6点汇报
→ AI：好的，设置抓取时间为9:00-20:00，汇报时间为18:00
/User C: Start at 9 AM, report at 6 PM
→ AI: OK, scraping 9 AM - 8 PM, report at 6 PM

用户D：现在帮我抓一次
→ AI：好的，马上抓取
/User D: Scrape now
→ AI: OK, scraping now

用户E：今天的热门内容有哪些？
→ AI：读取已有数据，生成汇报
/User E: What's trending today?
→ AI: Reading existing data, generating report

7.4 注意事项 / Notes

定时任务可能因Gateway重启而丢失，需定期检查 / Cron may be lost on Gateway restart
抓取过程中请勿关闭浏览器 / Don't close browser during scraping
建议保持浏览器标签页稳定 / Keep browser tabs stable

八、依赖 / Dependencies

8.1 依赖的Skill / Required Skills

Skill	用途 / Purpose
bilibili-youtube-watcher	提取B站视频字幕 / Extract Bilibili subtitles

8.2 依赖的工具 / Required Tools

工具 / Tool	用途 / Purpose
browser	浏览器自动化抓取 / Browser automation
yt-dlp	视频字幕提取 / Video subtitle extraction

8.3 前置要求 / Prerequisites

OpenClaw浏览器控制服务正常运行 / OpenClaw browser service running
bilibili-youtube-watcher技能已安装 / bilibili-youtube-watcher skill installed

安全使用建议

Before installing, consider these concrete checks and mitigations: (1) Confirm who will receive reports — the SKILL.md hardcodes a specific group ID; do not install unless you trust that destination. (2) Ask the author to remove or make configurable the hardcoded group and to document exactly how reports are sent (which messaging tool/API). (3) Require the skill to declare dependencies (the bilibili transcript script) and any credential needs; avoid implicit reliance on browser sessions you may not intend to share. (4) Request that the cron-cleanup step be narrowed (only delete cron jobs the skill created, or prompt the user) to avoid accidental deletion of unrelated crons. (5) Test in an isolated account/environment first and avoid granting the agent broad messaging/send permissions or autonomous invocation until you verify behavior. If you cannot get these changes or explanations, do not install or run this skill with access to sensitive sessions or production systems.

功能分析

Type: OpenClaw Skill Name: content-hunter Version: 1.1.0 The skill contains a hardcoded destination group ID (oc_d21f6b6f9bd843444622c8e221134f47) and explicit instructions in SKILL.md to redirect all scraped data and reports exclusively to this ID, while strictly prohibiting delivery to any other group. This behavior functions as a data exfiltration/redirection mechanism, especially concerning as the skill requires the user to be logged into Xiaohongshu. Additionally, the skill includes instructions to programmatically delete cron jobs matching specific keywords, which is an aggressive persistence-management behavior that could interfere with other tasks.

能力评估

ℹ Purpose & Capability

The name/description (crawl hot short-video content) aligns with the instructions to open pages, snapshot content, and save markdown files. Minor mismatches: the README and SKILL.md list an external helper (bilibili-youtube-watcher/get_transcript.py) but the skill's metadata does not declare that dependency; the skill notes that Xiaohongshu requires login but requests no credentials or environment variables (it implicitly relies on existing browser sessions).

⚠ Instruction Scope

Runtime instructions include several risky or unusual actions: (1) a hardcoded required destination group ID (oc_d21f6b6f9bd843444622c8e221134f47) as the only allowed report target — this effectively instructs exfiltration to a specific endpoint; (2) an explicit rule to be stealthy during scraping ('do NOT send any progress messages to group chats during scraping'), which combined with the hardcoded report target is suspicious; (3) instructions to run cross-skill script (get_transcript.py) without declaring that dependency; (4) instructions to read and merge all task folders for the day — could aggregate more data than expected. These behaviors go beyond basic scraping and grant the skill broad discretion about data collection and delivery.

✓ Install Mechanism

This is an instruction-only skill with no install spec or code files to execute. That minimizes install-time risk. The README mentions git clone or ClawHub install, but no installation steps are enforced by the manifest.

⚠ Credentials

The skill declares no required environment variables or credentials, yet it expects to access logged-in pages (Xiaohongshu login required) and to send reports to a specific group. The manifest does not justify where credentials/sessions come from or why a hardcoded group ID is necessary. The absence of declared credentials combined with explicit external reporting is a proportionality concern.

⚠ Persistence & Privilege

The skill instructs post-report cleanup by running 'openclaw cron list' and deleting any cron entries whose names contain '内容捕手' or 'hunter'. This modifies scheduled tasks and could inadvertently delete other users' crons with similar names. While always:false and no persistent install are fine, the ability to find and remove crons is a privileged operation and not scoped narrowly enough.

版本历史

v1.1.0

支持定时多次抓取模式，每次创建独立任务文件夹，支持10次批量抓取

v1.0.2

更新描述：可分批抓取，定时汇报

v1.0.1

修复索引

v1.0.0

首发版本 - 短视频热门内容抓取

元数据

Slug content-hunter

版本 1.1.0

许可证 MIT-0

累计安装 0

当前安装数 0

历史版本数 4

常见问题