wechat-article-fetcher

Pass

Audited by Gen Agent Trust Hub on Mar 11, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is susceptible to Indirect Prompt Injection (Category 8) because it processes untrusted content from WeChat articles. Ingestion points: scripts/fetch.py and wechat_article_fetcher.py extract text and images from external URLs. Boundary markers: The extracted content is returned without delimiters or instructions to ignore embedded commands. Capability inventory: subprocess.run is used across several files (scripts/run_in_venv.py, fetch_direct.py, wechat_article_fetcher.py) to execute internal scripts. Sanitization: While filename sanitization is performed, the extracted article text is not sanitized or escaped before being returned to the agent context.
  • [COMMAND_EXECUTION]: Several components of the skill utilize subprocess.run to manage script execution within isolated virtual environments. These commands are correctly constructed using argument lists, which prevents shell injection, but they allow the skill to execute internal Python logic based on user-provided parameters like URLs.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 11, 2026, 02:32 PM