atomic-commit
Pass
Audited by Gen Agent Trust Hub on Aug 25, 2026
Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes local shell commands including git status, git add, and git commit to manage the repository. These operations are essential for the skill's purpose and are constrained by a rule forbidding automatic remote pushing.
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data by reading file differences via git diff. 1. Ingestion points: File content retrieved through git diff as described in SKILL.md. 2. Boundary markers: Not explicitly defined for the diff content. 3. Capability inventory: Local file staging and committing (git add, git commit) defined in SKILL.md. 4. Sanitization: No sanitization of diff content is performed before processing for commit message generation. This surface is inherent to git operations and carries low risk due to lack of network or execution capabilities.
Audit Metadata