lov-article-creator
Pass
Audited by Gen Agent Trust Hub on Sep 1, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The profile management system in scripts/profile_store.py includes an explicit blacklist for sensitive data keys such as 'token', 'secret', and 'password', preventing accidental storage of credentials.
- [SAFE]: The article validation pipeline in scripts/validate_article_package.py automatically detects and flags the presence of private absolute user paths in output files, protecting user privacy.
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted research materials but successfully mitigates indirect injection risks through a 'reader contract' protocol and automated 'cold-reader' validation of article openings (implemented in skills/article-writing/SKILL.md and scripts/validate_article_package.py).
- [SAFE]: The skill uses standard, well-maintained Python libraries for image processing and data formatting within a strictly local functional scope.
Audit Metadata