receiving-code-review
Pass
Audited by Gen Agent Trust Hub on Nov 13, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and act upon code review feedback, which is external, untrusted data that could contain malicious instructions.
- Ingestion points: Code review comments and feedback received from both 'Human Partners' and 'External Reviewers' as described in SKILL.md.
- Boundary markers: The instructions do not define specific delimiters for the feedback content or include explicit warnings for the agent to ignore instructions embedded within the review text itself.
- Capability inventory: The skill utilizes the
grepcommand for codebase exploration and implies the use of file-writing or modification capabilities during the 'IMPLEMENT' phase. - Sanitization: Although the skill encourages questioning whether a suggestion 'breaks existing functionality', it lacks formal sanitization or validation of the feedback content to prevent the agent from following embedded prompt injection attempts.
Audit Metadata