caveman mode?

would love your thoughts after.

didn't it always read it?

It's interesting that I've also gone down a similar journey and the same keywords, fences, gates .. also showed up in my ai government platform.

yea - that's exactly what i saw. I like claude code, but err.... I wish they allowed better auditability. I'm using headroom, so i see somethings but I can't relay it back to what anthropic has.

2 factor vote. Vote in app, still go in person to validate result.

Here's me playing around with it. https://www.youtube.com/watch?v=2oJohBiqMUA

My approach/prototype using XState with websockets from an MCP server https://github.com/uptownhr/mcp-agentic-ui

appreciate it.

On why I built it and how. https://uptownhr.com/blog/building-lattice-knowledge-graph-c...