The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to create and execute shell scripts (e.g., scripts/run-component-lint.sh) to automate project quality gates. It also involves running local vendor scripts such as plugins/attune/scripts/attune_init.py for initialization.
[EXTERNAL_DOWNLOADS]: The configuration templates include references to well-known GitHub repositories for pre-commit hooks, including pre-commit/pre-commit-hooks, astral-sh/ruff-pre-commit, and PyCQA/bandit. These are recognized as safe sources for development workflows.
[PRIVILEGE_ESCALATION]: The skill uses chmod +x to make generated quality-check scripts executable. This is a standard and necessary step for the skill's primary purpose of setting up local automation.
[INDIRECT_PROMPT_INJECTION]: The skill defines scripts that ingest external data from git diff to determine which components to test. While this represents a potential attack surface, the scripts include directory existence checks ([ -d "plugins/$component" ]) and the usage is standard for monorepo development.
Ingestion points: git diff --cached --name-only in modules/component-level-hooks.md.
Boundary markers: Absent.
Capability inventory: Shell execution of make, uv run, ruff, mypy, and pytest through generated scripts.
Sanitization: Verification that the identified component corresponds to an existing directory.

precommit-setup