git-commit

Pass

Audited by Gen Agent Trust Hub on Mar 2, 2026

Risk Level: SAFE
Full Analysis
  • [PROMPT_INJECTION]: The skill contains instructions for branch naming and commit message formatting. These are legitimate task-oriented constraints and do not attempt to bypass AI safety filters or override system instructions.\n- [COMMAND_EXECUTION]: The skill utilizes standard Git CLI tools (git status, git commit, git log) to manage local repository changes, which is the primary and intended purpose of the skill.\n- [INDIRECT_PROMPT_INJECTION]: The skill processes repository data, which is a standard surface for indirect prompt injection inherent to git automation tasks.\n
  • Ingestion points: Repository file status and change analysis via git status in SKILL.md.\n
  • Boundary markers: None present to isolate repository data from the instruction context.\n
  • Capability inventory: Local git operations including git commit, git status, and git log.\n
  • Sanitization: Not applicable as the data is used for generating descriptive commit messages within the local repository environment.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 2, 2026, 10:37 AM