The Agent Skills Directory

[COMMAND_EXECUTION]: The skill facilitates the execution of shell commands through the prek CLI, including subcommands like prek run, prek install, and prek validate-config.
[COMMAND_EXECUTION]: The documentation provides examples of 'system' hooks that execute arbitrary local commands (e.g., just lint-md) as part of the pre-commit workflow.
[EXTERNAL_DOWNLOADS]: The skill refers to external sources for tool installation, including official package registries (npm, cargo) and GitHub repositories (j178/prek-action).
[INDIRECT_PROMPT_INJECTION]: The skill operates on repository configuration files (prek.toml, .pre-commit-config.yaml).
Ingestion points: Reads hook definitions from local configuration files.
Boundary markers: None present; the agent is instructed to trust and validate these files.
Capability inventory: The tool can execute any command defined in the configuration's entry field via prek run.
Sanitization: None; the skill relies on the user or the prek validate-config command to ensure configuration integrity.

prek