git-commit-expert

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFE
Full Analysis
  • Data Exposure & Exfiltration (SAFE): The skill includes non-negotiable security protocols that explicitly prevent the commitment of secrets, credentials, and environment files (.env). It does not contain any network exfiltration patterns.
  • Command Execution (SAFE): The skill executes standard development tools (git, npm, pytest, cargo) for the purpose of code verification. It explicitly forbids dangerous flags such as --force, --hard, or --no-verify unless specifically requested by the user, adhering to the principle of least privilege.
  • Prompt Injection (SAFE): No malicious override or bypass instructions were detected. The skill instructions reinforce agent safety and professional conduct.
  • Indirect Prompt Injection (LOW):
  • Ingestion points: Reads project files (package.json, README.md, Makefile) and repository history (git log, git diff).
  • Boundary markers: Employs a 'Decision Protocol' and 'Interaction Strategy' to maintain logical control.
  • Capability inventory: Execution of local build/test toolchains and git management.
  • Sanitization: Standard for development agents; the risk is mitigated by the safety-first commit protocols.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 06:06 PM