review-hog-resolution-criteria
Pass
Audited by Gen Agent Trust Hub on Oct 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data from pull request review threads and code repositories, which serves as a potential vector for indirect prompt injection.
- Ingestion points: Pull request threads (all comments) and the codebase at current head (SKILL.md).
- Boundary markers: There are no explicit delimiters or instructions provided to the agent to distinguish between system instructions and data ingested from PR comments.
- Capability inventory: The skill possesses the capability to modify code and commit changes to a repository (implementing 'fixed' outcomes).
- Sanitization: No sanitization or validation of the content within the review threads is performed before processing.
- Risk: Maliciously crafted comments, particularly those using trigger phrases like "SAFE TO FIX," could be used to override the agent's internal reasoning and trick it into committing harmful or unauthorized code changes.
Audit Metadata