The Agent Skills Directory

[COMMAND_EXECUTION]: The skill authors bash scripts, modifies file permissions via chmod +x, and validates script syntax using bash -n and shellcheck.
[DATA_EXFILTRATION]: The skill collects user-provided secrets (e.g., API keys) and stores them in local .env files or GitHub Secrets. It uses read -rs for secret entry to ensure credentials are not visible in terminal history.
[EXTERNAL_DOWNLOADS]: The generated scripts interface with GitHub via the gh command-line tool to manage repository secrets and variables.
[SAFE]: The skill emphasizes human-in-the-loop verification, requiring users to confirm the procedure scope and individual steps before execution.

wizard