The Agent Skills Directory

[SAFE]: The skill provides a comprehensive design toolset with robust, built-in security controls that align with best practices.
[SAFE]: The local iteration server (live-server.mjs) is secured with a randomly generated UUID token and restricted to 127.0.0.1, effectively preventing unauthorized external access.
[SAFE]: Automated scanning logic in the detector engine (hook-lib.mjs) includes a non-configurable skip list for sensitive files like credentials, private keys, and environment variables.
[SAFE]: High-privilege operations, including file writes and the spawning of specialized helper agents for asset production, are executed only in response to explicit user triggers within the design workflow.
[SAFE]: External network communication is used for legitimate tool functions, such as fetching configuration from the official vendor domain and interacting with well-known AI service providers (e.g., OpenAI).

impeccable