NYC

mastering-git-cli

Pass

Audited by Gen Agent Trust Hub on Feb 16, 2026

Risk Level: LOW
Full Analysis
  • COMMAND_EXECUTION (LOW): The scripts cleanup-agent-worktrees.sh, git-health-check.sh, setup-agent-worktrees.sh, and submodule-report.sh execute standard Git plumbing and porcelain commands. These are legitimate operations for the skill's stated purpose of managing repository state and worktrees.
  • INDIRECT_PROMPT_INJECTION (LOW): The skill handles repository metadata (e.g., branch names, submodule paths). While these could theoretically contain malicious content in an adversarial repository, the scripts perform read-only reporting or standard Git operations that do not execute the content of the repository data as commands or instructions to the LLM.
  • DATA_EXPOSURE (SAFE): The git-health-check.sh script inspects the repository state but does not access sensitive files outside the .git directory or user-specified paths. No network operations were detected that would exfiltrate this data.
Audit Metadata
Risk Level
LOW
Analyzed
Feb 16, 2026, 09:02 AM