git-pr-review

Pass

Audited by Gen Agent Trust Hub on Jun 22, 2026

Risk Level: SAFE
Full Analysis
  • [PROMPT_INJECTION]: Evaluated the use of directive language related to ignoring instructions. These phrases are part of a defensive strategy designed to protect the agent from following malicious commands found in external commit messages rather than being an attempt to bypass safety filters.
  • [COMMAND_EXECUTION]: The skill executes standard git commands to retrieve commit metadata and diffs. These operations are limited to reading repository information and do not involve administrative privileges or network access.
  • [SAFE]: The skill identifies commit messages and diffs as untrusted input and provides explicit instructions to treat them as inert data. This defensive layering is a best practice for agents processing external content.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 22, 2026, 12:14 PM
Security Audit — agent-trust-hub — git-pr-review