receiving-code-review
Pass
Audited by Gen Agent Trust Hub on Sep 23, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process feedback from external, untrusted reviewers (e.g., GitHub pull request comments).
- Ingestion points: Feedback from 'External Reviewers' as described in the Source-Specific Handling section of SKILL.md.
- Boundary markers: The skill mandates a specific 'Response Pattern' and a list of verification checks (e.g., 'Check: Technically correct for THIS codebase?') to serve as logical boundaries.
- Capability inventory: Instructions include using grep to search the codebase and gh api to interact with GitHub pull requests.
- Sanitization: The skill lacks technical input sanitization, relying instead on the agent's reasoning and manual verification steps.
- [COMMAND_EXECUTION]: The skill instructs the agent to use command-line tools to verify implementations and communicate feedback.
- Evidence: Mentions the use of grep for checking feature usage ('grep codebase for actual usage') and gh api for replying to comment threads ('gh api repos/{owner}/{repo}/pulls/{pr}/comments/{id}/replies').
Audit Metadata