Goal Decomposer

Big goals fail for one reason: the gap between where you are and where you want to be is so large that the next step is unclear. "Grow revenue to £500K" is inspiring but not actionable. "This week, send 15 outreach emails to fintech consultancies in London" is actionable. The decomposition bridges the two — annual target down through quarterly milestones, monthly checkpoints, and the specific weekly actions you actually need to do.

The skill starts with three pressure-tests. Specificity — what does success look like, what's the number, by when? Goals without deadlines are wishes. Reality — where are you starting, what's the gap, what resources do you have? Motivation — why does this matter and what happens if you don't hit it? These answers determine whether the plan is realistic or fantasy.

Then it finds the critical path: the 2-3 levers that drive most of the progress. Three questions surface them. "If you could only do ONE thing for the next 90 days, which activity moves you furthest?" That's the lever. "What's the one thing that, if you don't figure it out, makes the goal impossible?" That's the bottleneck. "What skill, asset, or relationship would accelerate everything else?" That's the multiplier. Decomposition without these is just a calendar of activity.

The plan layers four levels: annual goal, quarterly milestones (Q1 build foundation, Q2 accelerate, Q3 scale, Q4 close the gap), monthly checkpoints, and weekly actions — but only for the current month. Planning weekly actions for month eight is wasted effort because too much will change. The rule is detail for now, themes for later. Each quarterly milestone has a measurable indicator so "on track" isn't a feeling.

A few things the skill deliberately does. It works backward from the deadline, not forward from today — starting from the end reveals whether the timeline is realistic before you waste three months. It refuses to decompose vague goals ("be happier," "get in shape") and pushes for sharpening first. It explicitly considers killing the goal if decomposition reveals it's wrong, misaligned, or not actually what you want — that's a valuable outcome, not a failure. And every plan ends with the very first action, the thing to start now, so the user closes the session knowing exactly what to do Monday morning.