git-commit-trailers

Pass

Audited by Gen Agent Trust Hub on Apr 21, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: No security concerns were detected. The skill uses standard Git commands for repository metadata management and follows established development practices.
  • [INDIRECT_PROMPT_INJECTION]: The skill reads commit messages from the Git log to parse trailers. Although commit messages are untrusted data, the skill uses structured parsing tools like git interpret-trailers and filtering utilities like grep, sort, and uniq, which prevents the agent from executing instructions embedded in commit text.
  • Ingestion points: Commit logs via git log in SKILL.md
  • Boundary markers: Absent
  • Capability inventory: git log, git interpret-trailers, git config
  • Sanitization: Absent
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 21, 2026, 01:17 AM
Security Audit — agent-trust-hub — git-commit-trailers