review-and-ship

Pass

Audited by Gen Agent Trust Hub on May 22, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill utilizes standard command-line tools (git, gh) to perform intended development tasks such as fetching changes, checking status, and managing pull requests. These operations are performed within the context of the local repository and official GitHub services.
  • [DATA_EXPOSURE]: No sensitive file access (e.g., .ssh, .aws, or .env files) or hardcoded credentials were found. Data access is limited to the repository context for the purpose of code review.
  • [REMOTE_CODE_EXECUTION]: While the workflow involves running tests, this is a core function of a code review skill. No patterns of downloading and executing untrusted remote scripts (e.g., curl | bash) were detected.
  • [PROMPT_INJECTION]: The skill does not contain instructions attempting to override agent safety guardrails or bypass system constraints. It emphasizes correctness and security in its own guidelines.
Audit Metadata
Risk Level
SAFE
Analyzed
May 22, 2026, 06:45 AM
Security Audit — agent-trust-hub — review-and-ship