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, and gh API 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 the gh and git command-line tools to interact with the repository and GitHub services. \n- Evidence: Instructions include shell commands like gh pr view and procedures for creating checkouts and committing fixes.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 9, 2026, 10:36 PM
Security Audit — agent-trust-hub — respond-to-pr-review