The Agent Skills Directory

[SAFE]: The skill uses grep for local analysis and agentic_fetch for documentation retrieval, which are appropriate for a research tool. No malicious commands or suspicious behaviors were detected.\n- [PROMPT_INJECTION]: Instructions were reviewed for behavioral overrides, jailbreaks, or safety bypasses. The skill maintains a consistent focus on library auditing without attempting to subvert agent constraints.\n- [NO_CODE]: The skill provides markdown instructions and HTML templates. The only script included is a small, benign JavaScript snippet for navigation in the generated report.

library-deep-dive