git-commit-discipline
Pass
Audited by Gen Agent Trust Hub on Oct 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: No malicious patterns were detected. The skill promotes safe Git usage, including the protection of secrets and careful handling of destructive operations. It focuses on repository hygiene and recoverability checkpointing.\n- [INDIRECT_PROMPT_INJECTION]: The skill instructions involve the agent interacting with repository data, which creates a surface for indirect prompt injection.\n
- Ingestion points: The agent reads potentially untrusted data from the local repository using
git diff,git log,git status, andgit reflog(SKILL.md).\n - Boundary markers: The skill includes explicit instructions for the agent to manually inspect status and diffs, and to account for every staged line before committing, providing a layer of oversight.\n
- Capability inventory: The skill uses standard Git commands like
git add,git commit,git reset, andgit restoreto modify the repository and filesystem (SKILL.md).\n - Sanitization: The guidelines mandate that the agent never stage secrets or unrelated work and verify that each proposed checkpoint is coherent and verified.
Audit Metadata