i told claude to look in its stored conversation for how long a thing took previously
its letta subconscious already found the answer, quickly, despite not being installed when i did it
a good ten minutes and many subtool calls later, claude itself is still working on it via agents
night and day
Meet Claude's subconcious - an experimental plugin for Claude Code (built on @letta.com) that gives your Claude a split brain.
Claude's subconcious watches everything Claude does, forming memories in the background while whispering hidden messages back to "main Claude".
⭐ Letta @letta.com has just reached 20,000 stars on GitHub! Huge congrats! 🎉
Build and deploy AI agents with advanced memory
Letta Code is a big release for us, and it is a fundamentally different approach to coding agents with a focus on continual learning and statefulness.
One agent per project, specialized, infinitely lived.
No more compactions, no more forgetting.
Letta Code is also the #1 model-agnostic OSS harness on Terminal-Bench, with comparable performance to model-specific harnesses (Claude Code, Gemini CLI, Codex) across frontier models
We're releasing Letta Code, a memory-first coding agent
- open source (apache 2.0)
- model agnostic
- portable agent learning and memory
I tried this with the github MCP but it's horrible, so we might try again with zapier or something that can push to a letta agent
You can checkout the quickstart here docs.letta.com/quickstart
Let me know what's unclear - happy to answer any questions!
Stateful Agents meetup happening rn @ @cloudflare SF w/ Letta AI cofounders @sarahwooders.bsky.social @charlespacker.bsky.social as they build a stateful agent live on Letta
