The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes multiple shell commands to manage git worktrees (e.g., git worktree add, git worktree list, git worktree remove, git worktree prune), searches for configuration preferences using grep, and updates project files like .gitignore using echo.
[EXTERNAL_DOWNLOADS]: The skill triggers the installation of external dependencies using package managers like npm, pip, poetry, and go. These tools download software from well-known registries (e.g., npmjs.com, pypi.org), which is expected behavior for developer tools but constitutes fetching external code.
[PROMPT_INJECTION]: The skill is subject to indirect prompt injection vulnerabilities because it processes untrusted data from repository files that influence shell command execution and agent behavior.
Ingestion points: The skill reads directory preferences from CLAUDE.md and project requirements/scripts from manifest files such as package.json, Cargo.toml, and requirements.txt.
Boundary markers: No explicit delimiters or instructions are provided to the agent to treat the content of these files as untrusted or to ignore potentially malicious embedded instructions.
Capability inventory: The agent has the capability to execute shell commands, install packages, and run project-defined test suites (e.g., npm test).
Sanitization: No evidence of sanitization or validation is present to ensure that data retrieved from these project files does not contain malicious payloads targeting the shell or the agent's logic.

using-git-worktrees