多个微信公众号文章同步和下载

This skill appears to implement a full WeChat MP scraping/sync tool (login via QR, fetch article lists/details, download images, export Markdown, send progress notifications). Before installing or running it, consider: - Metadata omissions: the package requires Python packages (requirements.txt) and invokes an external 'openclaw' CLI, but the registry metadata lists no required binaries or install steps — you must install dependencies and ensure 'openclaw' is present yourself. - Local persistence: it will create directories, write a SQLite DB, save images and Markdown exports, and store WeChat login/session data on disk. If you don't want that data persisted on your machine, do not run it or run in an isolated environment. - Notifications/exfiltration risk: sync-all and other flows call 'openclaw message send' via subprocess to notify targets. Ensure any messaging targets/channel/account configured for notifications are trusted; otherwise the tool could forward scraped content externally. - Proxies and network access: the tool supports configuring proxies and will access WeChat endpoints; review network behavior (inspect mp_client.py and login_service.py) if you need to limit outbound connections. - Run in sandbox: because it executes subprocesses and writes files, run it in a controlled environment (VM or container) or inspect the code (notably mp_client.py, login_service.py, openclaw_messaging.py, and any code that performs HTTP requests) before use. If you want to proceed, install dependencies in a virtualenv, audit mp_client/login_service for endpoints and credential storage, and confirm 'openclaw' command behavior and where notifications would be sent.