skills/apocohq/skills/cleanup-commits/Gen Agent Trust Hub

cleanup-commits

Pass

Audited by Gen Agent Trust Hub on Mar 20, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses several git commands including git status, git log, git diff, git reset, git add, and git commit to manipulate the local repository state. These commands are fundamental to the skill's purpose of reorganizing commits and are executed within a clearly defined workflow.
  • [INDIRECT_PROMPT_INJECTION]: The skill analyzes external data in the form of existing commit messages and code diffs to help the agent formulate a commit plan. This creates a surface where malicious instructions embedded in a repository's history could potentially influence the agent's logic. However, the skill explicitly requires that the agent show the plan to the user and obtain confirmation before any destructive git operations are performed, providing a human-in-the-loop security check.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 20, 2026, 02:17 PM
Security Audit — agent-trust-hub — cleanup-commits