The Agent Skills Directory

[SAFE]: The skill facilitates the creation of interactive 'wizard' scripts for configuration tasks. While these scripts handle sensitive credentials (API keys, secrets), they do so following security best practices:\n
Hidden Secret Entry: The ask_secret function uses read -rs to prevent secrets from being echoed to the terminal screen during input.\n
Least Privilege and Persistence: The skill encourages writing secrets only to local .env files or using official tools like the GitHub CLI (gh secret set) for CI/CD environments. It does not perform unauthorized network exfiltration.\n
Safe URL Handling: The open_url function uses standard system commands (xdg-open, open, wslview) to open browser URLs provided by the script author or user.\n
Idempotency and Transparency: The library functions for environment management (write_env) use temporary files and standard stream redirection safely to prevent data loss or unauthorized access.

wizard