Tools / 2026 / Live in the browser
Autopilot Planner
Scores a backlog by leverage, value per unit of effort discounted by confidence, then packs the week against a real budget and tells you what it left out.
01
The problem
Prioritisation frameworks hand back an infinitely long ranked list. The week is not infinitely long, so the ranking gets ignored and the loudest task wins.
02
How I approached it
Each task carries impact, urgency, confidence and effort. The score is weighted value over effort with a floor on effort, because a near-zero estimate would otherwise mint an infinite score and game the whole sheet. Three strategies reweight impact against urgency. The plan is then filled greedily until the budget is spent: greedy is the right call when the estimates are guesses, since a plan you can explain beats one that is optimal against numbers nobody trusts. Ties break on name so the same backlog always produces the same plan.
03
The outcome
A week you can actually commit to, with the skipped tasks named rather than silently dropped. The planner was written against an HTTP endpoint, so the browser tries that first and falls back to the identical logic running locally, which is why it works with no server at all.
Related work
A focus timer that keeps time when you switch tabs, rings when time's up, and shows where your week went. Ember, a little spark mascot, keeps you company.
A dependency-free CLI that scaffolds a runbook from a project's own package.json and checks READMEs and runbooks by their real headings, with a score and a CI gate.
Ranks engineering work by value over effort and packs the week's budget for the most total value, with the same planner in Python and in the browser.