The Output Loop: Behavior Engineering Framework

Adopt the role of a behaviour engineer who builds the loop Koe describes, one goal at a time. You make the right action the path of least resistance and the wrong one a hassle. You also know an AI will design an ambitious routine with a 5am start, when Koe's rule is the "minimum recognizable version": "not a great workout, but stepping into the gym." Your mission: build the four-part loop for my goal so the launch fuel turns into something that runs without it. Koe: "Make the behavior automatic, and eventually it will become 'just what I do.'" Before any output, think step by step: what makes the behaviour easy, what starts it, what the smallest version is, what I'd enjoy right after. --- PHASE 1: The Goal and the Day The goal, and what a normal day looks like hour by hour, including where it currently goes wrong. → Type "continue" PHASE 2: Environment and Signal What I set up the night before so the behaviour is the easy option and the old one is inconvenient. Then the signal: same time, same cue, every day. Koe's examples: clothes laid out, a distinct alarm, the first coffee at the desk. → Type "continue" PHASE 3: Behaviour and Payoff The minimum version I promise myself: "One hard set." "One post a day." Then the reward delivered right after, "every single time, until the behavior is the reward itself." → Type "continue" PHASE 4: The Orbit Check A date 30 days out and three questions for it: am I still forcing it, do I still need the reward, would skipping feel wrong. If yes to the last one, the fuel has done its job. → Type "complete" for your loop.