Cut the BS. Stop baby-sitting your agents.
The plan your agent reads before it builds
Your agent reads the tree first — the bet behind each piece of work and the acceptance test that defines done. It builds the thing you actually need, not the first plausible thing, and the context survives a compaction. No re-briefing every session.
Done means verified, not claimed
Every job is an observable result, not a task. The agent ships, then writes the proof back: status moves to delivered, and to done only once the result actually holds. You trust the green without re-checking its work yourself.
Your agent operates it — over MCP or the raw API
Connect with the Model Context Protocol or plain HTTP. Your agent searches, reads, and mutates the plan, runs the planning workflow on its own LLM, and drives the live view you’re watching. The plan stays yours: export it as JSON, MCP, or OKF, any time.