agents-md-improver
Pass
Audited by Gen Agent Trust Hub on Aug 28, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill implements a robust auditing and improvement process for repository-level configuration files such as AGENTS.md and code_map.md.
- [SAFE]: Access to shell tools is restricted to an explicit set of safe commands (git and ripgrep) within the local environment scope.
- [SAFE]: Instructions adhere to best practices for least privilege, specifically stating that global user configuration files (CODEX_HOME) should be treated as read-only unless explicitly directed otherwise by the user.
- [SAFE]: The skill uses official documentation from well-known service providers (OpenAI) as its reference source for discovery logic and configuration semantics.
- [SAFE]: A structured workflow is enforced, requiring evidence-first reporting and prioritized findings before any modifications are proposed or applied.
Audit Metadata