The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes the pre-commit CLI and Bun's shell execution API ($ and spawn) to perform operations such as installing git hooks, running validations on staged files, and managing hook environments.
[EXTERNAL_DOWNLOADS]: Fetches configuration files and installation scripts from well-known repositories on GitHub, including the official pre-commit organization and terraform-linters.
[REMOTE_CODE_EXECUTION]: Implements standard installation patterns for development tools (e.g., the tflint installer) and uses the pre-commit try-repo command to verify hook configurations from remote repositories as part of testing workflows.
[PROMPT_INJECTION]: The skill possesses an indirect injection surface as it processes .pre-commit-config.yaml files to execute hooks. While this is the intended behavior of the framework, it represents a surface where an agent's actions could be influenced by instructions embedded in a project's configuration.
Ingestion points: .pre-commit-config.yaml (processed in PreCommitManager.ts and HookValidator.ts)
Boundary markers: Absent; the agent is instructed to parse and act upon the configuration content.
Capability inventory: Subprocess execution via Bun, file system writes, and git hook modification.
Sanitization: None; the logic assumes the configuration originates from a trusted project environment.
[SAFE]: All identified behaviors and external references are consistent with the skill's primary purpose of providing a development automation framework. No malicious patterns or obfuscation techniques were detected.

pre-commit