using-git-worktrees
Pass
Audited by Gen Agent Trust Hub on Oct 4, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses shell commands to interact with Git (rev-parse, worktree, check-ignore) and to execute build and test tools (npm, cargo, pip, go).
- [EXTERNAL_DOWNLOADS]: The skill automates the retrieval of project dependencies from standard package registries through established tools like npm install, pip install, poetry install, and go mod download.
- [INDIRECT_PROMPT_INJECTION]: The skill possesses an attack surface as it performs actions (dependency installation and test execution) based on the presence and content of files within the repository.
- Ingestion points: Ingests information from files including package.json, Cargo.toml, requirements.txt, pyproject.toml, and go.mod.
- Boundary markers: No specific boundary markers or delimiters are defined to isolate data from instructions in the ingested files.
- Capability inventory: Capabilities include file system manipulation, network access for package management, and arbitrary shell command execution for testing.
- Sanitization: The skill relies on the inherent security of the development tools and explicitly requests user consent before initiating the worktree creation process.
Audit Metadata