respond-to-pr-review
Pass
Audited by Gen Agent Trust Hub on Sep 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from external PR reviews and comments which may contain adversarial instructions. \n- Ingestion points: External review text and GitHub comments are retrieved in SKILL.md Step 1 using
gh pr view. \n- Boundary markers: The process requires a human to explicitly start the skill and review a numbered list of all items before any action is taken. \n- Capability inventory: The skill can perform file modifications,git commit,git push, andghAPI operations. \n- Sanitization: Relies on human arbitration for almost all items; however, an exemption allows automated fixing of factual errors in code comments. \n- [COMMAND_EXECUTION]: The skill utilizes theghandgitcommand-line tools to interact with the repository and GitHub services. \n- Evidence: Instructions include shell commands likegh pr viewand procedures for creating checkouts and committing fixes.
Audit Metadata