overview
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data sourced from git commit history, which could contain malicious instructions designed to influence the behavior of the analyzing subagents.
- Ingestion points: The skill reads commit messages and code diffs using
git logandgit show(Phase 1 and Phase 3). - Boundary markers: The prompt template for the subagents lacks explicit delimiters or instructions to treat commit data as untrusted content, increasing the risk that instructions embedded in a commit could be followed by the agent.
- Capability inventory: The skill possesses shell execution capabilities (
git,mkdir,basename) and file-write access to the user's home directory (~/overviews/). - Sanitization: There is no evidence of sanitization or filtering of the retrieved commit content before it is processed by the subagents.
Audit Metadata