Long agent runs fail in a boring way: not with a dramatic crash, but by quietly waiting for the next...
Why Long Claude Code and Codex Runs Stall, and Why I Built a Local Notch Companion
Long agent runs fail in a boring way: not with a dramatic crash, but by quietly waiting for the next...
History is for humans. The present source of truth is for agents. Why I moved my effort from CLAUDE.md and AGENTS.md to the repo itself — schema, tests, lint, exports, state machin...
We are uncovering better ways of building software with AI agents. Through this work we have come to value certain things over others.
I want to talk about something that keeps coming up in product conversations and almost always lands...