respond-pr-feedback
Pass
Audited by Gen Agent Trust Hub on May 26, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill makes extensive use of the GitHub CLI (
gh) andjqto perform repository operations, such as querying PR details, fetching review comments, and executing GraphQL mutations to resolve threads. This also includes writing temporary logic files to/tmpfor data processing. - [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface as it ingests and processes pull request comments authored by external reviewers.
- Ingestion points: External review comments are fetched from the GitHub API in Step 3a.
- Boundary markers: No explicit delimiters or instructions are used to separate untrusted comment content from the agent's logic during evaluation.
- Capability inventory: The skill has the capability to post replies to PR comments and resolve review threads via the
ghCLI API and GraphQL endpoints. - Sanitization: No explicit sanitization or filtering of the comment body is described before the agent evaluates the feedback.
Audit Metadata