agent-md-improver
Pass
Audited by Gen Agent Trust Hub on Aug 4, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No malicious behavior detected. The skill is a utility for managing documentation and instructions within a local repository environment.
- [COMMAND_EXECUTION]: The skill utilizes standard search tools like
rg(ripgrep) to identify relevant markdown files. This is a common and expected utility for file discovery tasks. - [PROMPT_INJECTION]: The skill is designed to process markdown files which could potentially contain indirect prompt injection. However, the skill explicitly directs the agent to verify all content against repository evidence (code, config, CI/CD) and preserve existing safety guardrails, which serves as a robust mitigation strategy.
- [DATA_EXPOSURE]: The skill reads repository configuration files to determine the project's risk profile (e.g., distinguishing between experimental and production environments). This data access is scoped to the local repository and is used to ensure that security-sensitive instructions are not inadvertently weakened.
Audit Metadata