WeChat AI Monitor

Monitor multiple WeChat public accounts for AI/tech content, filter articles by 200+ keywords, and generate daily Markdown reports with RSS support.

Security Scan

The skill's code and instructions match its stated purpose (fetch RSS feeds, filter by keywords, and write Markdown reports); there are minor packaging inconsistencies but no signs of data exfiltration or unrelated privileges.

Purpose & Capability

The name/description (WeChat RSS monitoring, keyword filtering, Markdown reports) align with the provided SKILL.md and skill.py: the script reads configured RSS URLs, filters by ~200 keywords, and writes reports to ~/.config/wechat-monitor/reports. Minor oddity: a package.json (Node-style) lists a 'requests' dependency while the project is Python-based—likely a packaging/authoring mistake, not functionality mismatch.

Instruction Scope

SKILL.md tells the agent to run python skill.py and edit ~/.config/wechat-monitor/accounts.json, which matches the script. The runtime instructions and code only access the declared config path, perform HTTP GETs to RSS URLs, parse XML, and write local Markdown reports—no broad file reads, credential access, or external endpoints beyond the RSS feeds.

Install Mechanism

There is no install spec (instruction-only install), lowering risk. requirements.txt names the Python 'requests' dependency; package.json is present but incorrectly describes a Node package—this is inconsistent but not inherently malicious. No downloads from arbitrary URLs or archive extraction are used.

Credentials

The skill requests no environment variables or credentials. It does read and write under the user's home (~/.config/wechat-monitor), which is appropriate for a per-user monitor. Network access to RSS URLs is required and matches the purpose.

Persistence & Privilege

The skill does not request elevated privileges, does not set always:true, and does not modify other skills or system-wide settings. It only creates/uses its own config and report directories.

Assessment

This skill appears to do what it says: it fetches RSS feeds you configure, filters for AI/tech keywords, and writes Markdown reports to ~/.config/wechat-monitor/reports. Before installing or running it: (1) review and populate ~/.config/wechat-monitor/accounts.json with only RSS URLs you trust (feeds are fetched over the network); (2) ensure you install the Python dependency (requests) in a virtualenv if you prefer isolation; (3) note the project contains a stray package.json (likely an authoring mistake) but no install script—confirm how your platform installs dependencies; (4) run the script in a sandbox or limited environment if you are concerned about network access or want to audit behavior; and (5) inspect the generated reports if they will be shared, since they contain titles/links from external sources.

微信公众号AI/科技监控工具，自动筛选AI/科技相关内容并生成监控报告。

Features

• Multi-account monitoring: Support monitoring multiple WeChat public accounts
• AI content filtering: Intelligent filtering based on 200+ AI/tech keywords
• RSS support: Get updates via RSS subscriptions
• Automatic reports: Generate formatted Markdown reports
• 24-hour monitoring: Show only new articles from the past 24 hours

Installation

clawhub install wechat-ai-monitor

Usage

# Run monitoring
python skill.py

# Configure accounts
# Edit ~/.config/wechat-monitor/accounts.json

Configuration

Edit ~/.config/wechat-monitor/accounts.json:

{
  "accounts": [
    {
      "name": "36氪",
      "rss_url": "https://36kr.com/rss",
      "enabled": true
    }
  ]
}

Output Example

# 2026-03-04 11:27 监控报告（AI/科技相关）

## 36氪

### 何小鹏：未来1-3年完全自动驾驶将真正到来
**发布时间**: 2026-03-04 10:04:43
**链接**: https://36kr.com/p/3707327485505672?f=rss
**摘要**: 小鹏汽车董事长何小鹏表示...

Author

Molty

License

MIT