ProofPlan DECISION ENGINE
Local-only

For the student staring at five urgent things

Stop guessing which task to start.
Decide it. Prove it. Print it.

ProofPlan is a glass-box decision engine. Compare your real options on evidence, watch it classify the choice as a two-way door (decide fast) or a one-way door (slow down), and walk away with a Decision Receipt — the chosen path, the first move for today, and an if-then fallback for the one thing most likely to break it.

Try a real one:
1

Frame the decision

One line a judge — or future-you — could understand in five seconds.

2

Compare the paths

Score each path, then attach the evidence that makes it believable — not a vibe.

The glass box

No black box. Here's exactly why that path wins.

Effort × Payoff map Bubble size = fit · numbers match the ranked paths → · top-left is the sweet spot.
Ranked paths
How the score is computed (it's editable)

A transparent weighted matrix — no AI guess. Each path earns: Impact ×30%, Confidence ×20%, Feasibility ×20% (effort vs. your energy), Reversibility ×15%, Evidence ×15%, then a deadline-urgency nudge. Change any input and the ranking updates live — you can see and challenge every number.

Plan checks
Do this today
    If-then fallbacks
    Decision journal saved in this browser

    Why this works

    Built on decision science, not productivity theater.

    Two-way vs. one-way doors

    ProofPlan classifies your choice by reversibility. Reversible "two-way door" calls should be made fast at ~70% certainty; irreversible "one-way doors" deserve a slower, evidence-heavy process.

    Bezos, Amazon shareholder letters

    The premortem

    Imagining a decision has already failed helps people identify ~30% more reasons it could go wrong than asking "what might go wrong." That's why step 3 assumes failure first.

    Premortem: Klein, HBR (2007) · 30% figure: Mitchell, Russo & Pennington (1989)

    If-then implementation intentions

    Pre-deciding "if X happens, then I'll do Y" roughly doubles follow-through. In one study, gym attendance jumped from 39% to 91% with a single if-then plan; a 94-study meta-analysis found a medium-to-large effect.

    Gollwitzer & Sheeran (2006); Milne, Orbell & Sheeran (2002)

    Glass-box weighting

    Forcing the trade-off weights to be explicit beats gut feel on multi-option choices — and unlike an AI suggestion, every number here is visible and editable.

    Weighted decision matrix · decision hygiene, Kahneman, Sibony & Sunstein, "Noise" (2021)