matsakis-ownership-mastery
Fail
Audited by Gen Agent Trust Hub on Mar 9, 2026
Risk Level: HIGHPROMPT_INJECTIONNO_CODE
Full Analysis
- [PROMPT_INJECTION]: The titles in
SKILL.mdandreferences/philosophy.mdcontain extensive sequences of zero-width characters (U+200B, U+200C, U+200D, U+2060). This hidden content encodes non-visible data (beginning with the string 'SK1L_'), which is a high-risk indicator often used to evade static analysis or store hidden instructions within the agent's context. Mitigations: Remove all non-printing characters from skill metadata and instructions. - [NO_CODE]: The skill consists exclusively of Markdown documentation and Rust code examples intended for the agent to follow. No executable scripts, binaries, or automated tasks were found in the provided files.
Recommendations
- AI detected serious security threats
Audit Metadata