The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute standard development commands such as npm install, composer install, and various formatting tools (pint, prettier, biome). These are within the expected scope of a formatting and linting setup skill.
[EXTERNAL_DOWNLOADS]: The skill uses well-known GitHub Actions (actions/checkout@v5, shivammathur/setup-php@v2) and standard package registries (npm/composer). These sources are recognized as safe within the context of software development automation.
[REMOTE_CODE_EXECUTION]: While the skill involves running package managers that execute scripts during installation (e.g., husky's prepare script), these are standard practices for the tools involved. No arbitrary remote script execution or piped curl-to-bash patterns were found.
[DATA_EXFILTRATION]: There are no patterns suggesting the exfiltration of sensitive data. The network operations are limited to standard dependency management and CI environment setup.
[PROMPT_INJECTION]: The instructions are clearly technical and do not attempt to override the AI agent's safety guidelines or persona. The language used is descriptive and procedural.
[DYNAMIC_EXECUTION]: The runbook includes a small inline Python snippet in references/runbook.md used to programmatically update biome.json. This is a targeted, local configuration update using standard libraries and does not process untrusted external input.

formatting-enforcement