refactor-maintainer
Pass
Audited by Gen Agent Trust Hub on Sep 1, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes external code and repository metadata from the working tree, which serves as an ingestion point for untrusted data that could contain malicious instructions.
- Ingestion points: The skill instructs the agent to "inspect the working tree" and "find all callers" within the target repository.
- Boundary markers: The instructions lack explicit boundary markers or warnings to the AI to ignore instructions embedded within the processed code files.
- Capability inventory: The skill's workflow implies the use of tools to read and write to the file system ("migrate internal callers", "deletion of obsolete paths") and execute commands ("run focused checks", "repository's required final gate").
- Sanitization: No sanitization or validation of the ingested content is specified before it is processed by the agent.
Audit Metadata