governed-agent
by Jeremy
Capability-based security, scope-bounded planning and out-of-band audit for autonomous LLM agents.
A meta-agent built on the Claude Agent SDK that designs, writes and tests new agents through conversation.

Claude Agent SDK Playground holds Agent Builder, a command line tool that builds other agents. You describe the agent you want, it asks clarifying questions, designs its tools, writes its identity files and produces a runnable Python project.
Each agent it makes is defined by a set of markdown files covering its operating manual, personality and memory. The builder tests every new agent against mock tool outputs before handing it over, so it can be checked without live API calls or real side effects. It is for developers exploring the Claude Agent SDK and identity-driven agents.
The builder can edit files and run shell commands. The README has a safety note; run it in a folder you are happy for it to change.
by Jeremy
Capability-based security, scope-bounded planning and out-of-band audit for autonomous LLM agents.
by Chan
An open-source language for floor plans, now also inside ChatGPT: ask for a plan in plain words and ChatGPT writes it in ArchLang, compiles it and checks it.
by Ming
Copy on one device and paste on another, or share with people and agents, through a short-lived room or the rclip CLI.