The Agent Skills Directory

[SAFE]: The skill implements a mandatory human-in-the-loop approval gate using a instruction that prevents any code implementation or project scaffolding until a design has been explicitly approved by the user.
[SAFE]: References to external research tools such as 'octocode MCP' are presented as optional enhancements for gathering context and do not involve the automated installation or execution of untrusted code.
[SAFE]: The process follows secure development best practices by ensuring architectural decisions are documented, reviewed, and versioned before any development activity begins.

vs-brainstorm