using-git-worktrees
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill reads from local project files to determine worktree directory preferences, presenting an attack surface where maliciously crafted project files could influence agent behavior. 1. Ingestion points: Uses grep to extract 'worktree' preferences from CLAUDE.md. 2. Boundary markers: Lacks explicit delimiters or safety warnings for the ingested data. 3. Capability inventory: The skill can execute various shell commands including git operations, file modification, and standard package managers. 4. Sanitization: No sanitization or validation of the extracted configuration strings is performed.
- [COMMAND_EXECUTION]: The skill automatically detects project manifest files and executes corresponding build, dependency management, and test tools. 1. Evidence: Automates execution of npm install, cargo build, pip install, poetry install, and go mod download based on the presence of package.json, Cargo.toml, or requirements.txt. 2. Evidence: Runs project test suites (npm test, cargo test, pytest, go test) to verify the new worktree baseline state.
Audit Metadata