commit

Pass

Audited by Gen Agent Trust Hub on Mar 27, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill performs local git operations including git status, git diff, git add -A, and git commit -F. These commands are used to stage changes and create commits as part of the skill's primary function.\n- [PROMPT_INJECTION]: The skill exhibits a surface for indirect prompt injection as it processes untrusted data from the session history and repository diffs to generate commit messages and determine which files to stage.\n
  • Ingestion points: Reads session history for intent/rationale and git diff for change details in SKILL.md.\n
  • Boundary markers: Lacks explicit delimiters or 'ignore' instructions when processing external data into the commit generation prompt.\n
  • Capability inventory: Authorized to stage all changes via git add -A and commit them to the repository in SKILL.md.\n
  • Sanitization: Employs manual checks by instructing the agent to 'confirm scope' and 'sanity-check' files for artifacts or logs before proceeding, which serves as a mitigation against automated malicious staging.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 27, 2026, 03:29 PM