git-commit
Pass
Audited by Gen Agent Trust Hub on Sep 25, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [EXTERNAL_DOWNLOADS]: The README.md file suggests installation methods via npx and git clone from the author's own repository (walterra/agent-tools). These are standard deployment practices for the vendor and target trusted infrastructure.\n- [COMMAND_EXECUTION]: The skill uses git commands for diffing and committing, and invokes local project tools like npm or pnpm to fix linting issues as part of its documented workflow in SKILL.md. These commands are necessary for the skill's stated utility.\n- [INDIRECT_PROMPT_INJECTION]:\n
- Ingestion points: The skill analyzes file changes using git diff --cached in SKILL.md.\n
- Boundary markers: The instructions do not specify delimiters to separate the diff content from the agent's logic.\n
- Capability inventory: The skill executes shell commands (git, npm, pnpm) and can modify files via linting fixes and staging commands.\n
- Sanitization: No explicit sanitization of the diff output is performed before analysis.
Audit Metadata