The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes shell commands using the detected package manager to install dependencies and initialize the Husky environment (e.g., npx husky init).
[PERSISTENCE]: The skill modifies the repository's git hooks by creating a .husky/pre-commit file. This ensures that linting and testing scripts run automatically whenever a commit is made. This behavior is the primary stated purpose of the skill.
[EXTERNAL_DOWNLOADS]: The skill downloads and installs widely-used development tools (husky, lint-staged, prettier) from the official npm registry.
[INDIRECT_PROMPT_INJECTION]: The skill adapts its execution logic based on the contents of the repository's configuration files.
Ingestion points: Reads package.json and lock files (package-lock.json, pnpm-lock.yaml, yarn.lock, bun.lockb) to identify the project's environment.
Boundary markers: Not applicable for this specific workflow.
Capability inventory: File system write access for configuration, package installation via shell, and git commit execution.
Sanitization: The skill adapts to existing project scripts (typecheck/test) but does not perform explicit sanitization of the script names found in package.json.

setup-pre-commit