Priority Matrix

The default human approach to a long to-do list is to work the loudest item first — whoever messaged most recently, whatever's most urgent in someone else's head. That's reactive triage, and it produces weeks where you finish 30 tasks and none of them mattered. This skill replaces that with the ICE-D model — Impact, Cost of delay, Effort, Dependency — and surfaces what actually moves the needle versus what just feels urgent.

The scoring isn't decorative. Cost of delay is the dimension most lists ignore — a task that gets worse every day it waits (a deadline, a blocked person) outranks a task with no deterioration. Dependency catches the items that look small but unblock others. Effort is inverted, so the high-impact 20-minute task beats the high-impact 3-day task for today's slot. Items get sorted into Do Now / Schedule / Delegate / Eliminate, with the Eliminate category treated as a first-class output rather than a footnote — choosing not to do something is one of the most underused productivity moves and it gets named explicitly with reasons.

There's a hard cap of 3 items in Do Now. If everything scores into Do Now, the scoring needs recalibration — the matrix forces you to make a real choice rather than promote everything to "urgent." It also handles the patterns that look like prioritisation problems but aren't: a single overwhelming task you keep avoiding is a decomposition problem, not a ranking problem, and gets sent to task breakdown instead. A team backlog gets a resource-constraint dimension because N people means at most N items can be Do Now simultaneously.

For the moments when you don't have time for the full matrix, the two-question quick mode breaks paralysis fast: "What if not done today causes a real problem?" and "What will make everything else easier?" That's not as rigorous as the full scoring, but it's enough to start moving. The skill is opinionated about one thing — your top 3 is your real list. Everything below #3 is aspirational until those are done.