retro
Pass
Audited by Gen Agent Trust Hub on Oct 29, 2026
Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
- [SAFE]: The skill provides a methodical approach to improving agent efficiency through repository maintenance and instruction refinement.
- [COMMAND_EXECUTION]: The skill directs the agent to execute and verify local scripts, hooks, and tests to ensure that proposed improvements are functional within the repository environment.
- [DATA_EXPOSURE]: The skill includes an explicit safety protocol requiring the agent to sanitize all drafts intended for external repositories by removing sensitive information such as names, hostnames, and credentials.
Audit Metadata