group-commit
Pass
Audited by Gen Agent Trust Hub on Sep 23, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill makes extensive use of git commands, including 'status', 'reset', 'rev-parse', 'update-index', 'commit', and 'worktree', to manage the repository index and working tree. It also executes a bundled Node.js script ('scripts/hunks.mjs') for fine-grained hunk manipulation. These operations are core to the skill's functionality as a git management tool.
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to analyze uncommitted changes, which can contain untrusted data or malicious instructions embedded in code comments or documentation. \n
- Ingestion points: Working tree diffs and file contents are ingested via 'git diff' and the 'hunks.mjs' listing command in 'SKILL.md'. \n
- Boundary markers: No specific boundary markers or safety instructions are used to separate the code content from the analysis prompts. \n
- Capability inventory: The skill possesses the ability to execute shell commands via 'git', modify the git index, and perform build checks. \n
- Sanitization: The skill does not perform sanitization of the processed code changes before the agent evaluates them for commit grouping.
Audit Metadata