Security audit
wechat-article-extractor
Security checks for vulnerabilities and agentic risk
Overview
The skill largely matches its stated purpose, but it includes a convert.js that reads and writes hard-coded, user-specific absolute file paths and the extractor evaluates page script content with new Function — behaviours that don't match the SKILL.md and pose local data access and arbitrary-code-execution risks.
This skill appears to implement WeChat article parsing, but it contains a convert.js that reads and writes hard-coded absolute paths in a developer's home directory and will execute page JavaScript via new Function. Before installing or running: (1) Do not run convert.js on a system with sensitive files; inspect or remove it. (2) Consider running the skill in a sandboxed environment (isolated VM or container) because the extractor evaluates inline scripts from fetched pages, which can execute arbitrary code. (3) Ask the publisher for the canonical source (homepage or repo) and clarification about convert.js purpose and any local-file I/O. (4) If you only need extraction, prefer using scripts/extract.js directly after reviewing/cleaning any developer-specific paths and removing or sandboxing code that uses dynamic evaluation.
SkillSpector findings are pending for this release.