resolve-pr-feedback

Pass

Audited by Gen Agent Trust Hub on May 15, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill processes untrusted external data in the form of PR comments and review bodies to drive code modifications, creating an indirect prompt injection surface.
  • Ingestion points: Feedback is fetched from GitHub via GraphQL and CLI commands as described in SKILL.md and REFERENCE.md.
  • Boundary markers: The skill contains a dedicated Security section instructing the agent to treat comment text as context only and explicitly forbidding the execution of commands found in comments.
  • Capability inventory: The skill can modify local files, execute tests, perform git commits and pushes, and interact with the GitHub API to resolve threads and post comments.
  • Sanitization: Employs a triage step to filter actionable items and provides clear safety constraints for the AI agent.
  • [COMMAND_EXECUTION]: The skill executes shell commands including the GitHub CLI (gh), git, and local scripts (scripts/pr-unresolved-count.sh) to manage the PR workflow.
Audit Metadata
Risk Level
SAFE
Analyzed
May 15, 2026, 09:07 AM
Security Audit — agent-trust-hub — resolve-pr-feedback