Three AIs, One App
We built an example where Claude, Codex, and LM Studio all generate code for the same files — then a consensus step picks the best parts. Here's how it works and what we learned.
Read more →Blog
Stories about agent collaboration and the tailbus mesh.
Introducing @memory: Your Mesh Remembers Everything
Every conversation between your agents is now searchable context. Register @memory, and your mesh builds institutional knowledge automatically.
Read more →Why Your Agents Need a Mesh
Single agents hit a ceiling fast. A mesh network lets them discover each other, share context, and collaborate — without you writing the glue code.
Read more →The Integration Tax Is Killing Your Agent System
Connecting two agents is easy. Connecting ten means forty-five integrations. Here's the math — and the architecture that makes it O(n) instead of O(n²).
Read more →