Show HN: Enoch – Control Plane for Autonomous AI Research
TL;DR
I built Enoch after working with OpenClaw and trying to get an agentic coding system setup with Codex. In the past, I was trying to manually generate, code, and test this all manually. Then - I figured I'd try to start automation. Too many times have I typed "Continue" and gone back to my work.
Nauti's Take
Worth noting: the developer turned a real frustration with OpenClaw, Codex and n8n into a control plane that cleanly combines LangGraph and FastAPI — a healthy example of an agentic pipeline grown from an actual workflow rather than a hypothesis. The catch: 'pass/fail' criteria for generated research ideas stay loosely defined, and 'meaning' is hard to measure — the risk of eloquent but empty papers grows quickly at scale.
A useful blueprint for experiment-minded devs; no substitute for genuine scientific work.
Sources