wechat-article-fetcher

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSDATA_EXFILTRATION
Full Analysis
  • [Indirect Prompt Injection] (LOW): The skill is designed to ingest and process untrusted data from external WeChat URLs, creating a surface for embedded malicious instructions.
  • Ingestion points: External article content fetched from mp.weixin.qq.com via bin/main.py and utils/downloader.py.
  • Boundary markers: Absent; no mention of delimiters or instructions to the agent to ignore content within the fetched articles.
  • Capability inventory: Network GET/POST requests, local file system writes (images/ directory), and summarization logic.
  • Sanitization: No evidence of content sanitization or filtering of the fetched HTML/Markdown before it is processed by the agent.
  • [Data Exfiltration] (LOW): The skill facilitates sending data (article summaries) to an arbitrary external URL provided by the user as a 'flomo secret'. While intended for integration, this mechanism allows the agent to transmit processed information to any network endpoint.
  • [External Downloads] (SAFE): The skill performs automated downloads of image resources from WeChat servers. This is consistent with its primary purpose of article localization and does not involve executing remote code.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 05:33 PM