using-git-worktrees

Pass

Audited by Gen Agent Trust Hub on May 13, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes Git commands (git worktree add, git check-ignore, git rev-parse) to manage isolated workspaces. It also attempts to discover and execute project-specific setup and test commands from the repository's documentation or configuration files. This is standard behavior for a development-oriented AI agent skill.
  • [PROMPT_INJECTION]: The skill has a surface for indirect prompt injection because it reads and follows instructions from repository-controlled files such as AGENTS.md and README.md to determine how to set up and test the workspace.
  • Ingestion points: AGENTS.md, README.md, Makefile, package.json (within the target repository).
  • Boundary markers: Absent. The skill does not explicitly warn the agent to ignore instructions embedded in these files.
  • Capability inventory: Shell command execution via Git and platform-specific tooling.
  • Sanitization: Absent. However, the skill includes a safety step to report test failures to the user and ask for guidance before proceeding, which acts as a manual checkpoint.
Audit Metadata
Risk Level
SAFE
Analyzed
May 13, 2026, 03:36 PM