wechat-article-writer

Pass

Audited by Gen Agent Trust Hub on Aug 4, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill implements strong security boundaries and follows best practices for AI agent tools.
  • [COMMAND_EXECUTION]: Local Python scripts (scripts/render_wechat_layout.py and scripts/validate_image_manifest.py) are used for article validation and HTML rendering. These scripts are provided within the skill package and perform deterministic transformations without unsafe system calls or network access.
  • [PROMPT_INJECTION]: No malicious instruction overrides or bypass attempts were found. The prompt instructions are focused on legitimate style adaptation and content research.
  • [DATA_EXFILTRATION]: The skill documentation and pipeline contract explicitly forbid requesting credentials, cookies, or tokens, and strictly prohibit automatic publishing to external platforms.
  • [EXTERNAL_DOWNLOADS]: Research is conducted using standard agent capabilities (web search/fetch). There are no patterns of downloading and executing external code or binaries.
  • [INDIRECT_PROMPT_INJECTION]: The skill handles untrusted data from web searches but mitigates injection risks by using a restricted layout renderer that escapes all raw HTML and only supports a subset of safe Markdown elements.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 4, 2026, 06:28 AM
Security Audit — agent-trust-hub — wechat-article-writer