using-git-worktrees

Pass

Audited by Gen Agent Trust Hub on May 13, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructs the agent to execute several shell commands to manage the Git environment and project lifecycle, including git worktree, git check-ignore, and build-related tools such as npm, cargo, pip, and go.
  • [EXTERNAL_DOWNLOADS]: Automated dependency installation via npm install, pip install, and other package managers may trigger downloads from official public registries (such as npmjs.org or pypi.org) based on the contents of the project's configuration files.
  • [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface by ingesting data from the project's CLAUDE.md file to determine directory preferences.
  • Ingestion points: CLAUDE.md is searched via grep for worktree configuration strings.
  • Boundary markers: Absent; the skill does not explicitly define delimiters or instructions to ignore embedded commands within the ingested file.
  • Capability inventory: The agent is authorized to create directories, modify the local .gitignore file, install software packages, and execute project tests.
  • Sanitization: There is no explicit sanitization of the strings retrieved from CLAUDE.md before they are used in the directory selection logic.
Audit Metadata
Risk Level
SAFE
Analyzed
May 13, 2026, 07:15 PM
Security Audit — agent-trust-hub — using-git-worktrees