pr-feedback

Pass

Audited by Gen Agent Trust Hub on Jul 24, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses git and the gh CLI to interact with the repository and GitHub. These commands are used to identify the PR, manage branches, and post replies to review threads.
  • [PROMPT_INJECTION]: The skill has an indirect prompt injection surface as it ingests untrusted data from GitHub PR comments. This is mitigated by requiring manual user approval for all actions.
  • Ingestion points: SKILL.md (Step 2) fetches review threads and comments via gh api graphql and gh pr view.
  • Boundary markers: No explicit delimiters are used to separate instructions from comment content.
  • Capability inventory: The skill can write to the local file system (Step 5), execute shell commands (git, gh), and interact with the GitHub API (Step 7).
  • Sanitization: No sanitization is performed on the ingested comment data.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 24, 2026, 05:29 AM
Security Audit — agent-trust-hub — pr-feedback