skills/junerdd/skills/split-commits/Gen Agent Trust Hub

split-commits

Pass

Audited by Gen Agent Trust Hub on Sep 12, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill manages local Git state through standard commands such as git add and git commit. It includes strict instructions to avoid destructive operations and prohibits pushing changes to remote repositories.
  • [INDIRECT_PROMPT_INJECTION]: The skill creates a surface for potential indirect instruction injection by reading local file content via git diff to generate commit messages and plan commit batches.
  • Ingestion points: Working tree changes and diffs are read in SKILL.md via git diff and git status.
  • Boundary markers: There are no explicit delimiters or specific 'ignore' instructions for the content of the diffs themselves.
  • Capability inventory: The skill can stage files and create local commits.
  • Sanitization: The risks are mitigated by mandatory human review of all generated commit messages and the requirement for explicit user authorization for each commit batch.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 12, 2026, 04:30 AM