Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Build a debt payoff plan — avalanche vs snowball simulated month by month on your actual debts, the real payoff dates, and the psychology-vs-arithmetic tradeoff priced in dollars. Use when asked how do I pay off my debts, avalanche or snowball, make me a debt payoff plan, or when will I be debt-free. Produces the month-by-month comparison from the script, the payoff order with dates, the interest cost of choosing morale over math, and the plan-survival rules.
.claude/skills/mohitagw15856-debt-payoff/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-02 | ✗→✓ | ▲ Improved | 59% | 0% |
| case-07 | ✗→✓ | ▲ Improved | 110% | 0% |
| case-10 | ✗→✓ | ▲ Improved | 22% | 0% |
| case-22 | ✗→✓ | ▲ Improved | 80% | 0% |
| case-01 | ✓→✗ | ▼ Worse | -50% | 0% |
The avalanche-vs-snowball debate has a correct answer (avalanche, arithmetically) and a true objection (plans people abandon save nobody anything) — and the only honest way to weigh them is to price the difference on the actual debts. Sometimes morale costs $40; sometimes it costs $4,000. This skill runs both strategies month by month with the script, shows what choosing early wins actually costs, and builds the plan around the constraint that matters most: it has to survive contact with a real budget.
Ask for these if not provided:
bashpython3 scripts/debt_payoff.py --debt "visa:9000:24.9:180" --debt "loan:3000:6:60" --extra 200 python3 scripts/debt_payoff.py --debt "a:2000:19:50" --debt "b:9000:6:180" --extra 150 --json
Deterministic month-by-month simulation: interest accrues, minimums are paid, the surplus cascades onto the strategy's target debt, and freed-up minimums roll forward. Fixed APRs, on-time payments, no new charges — the assumptions are printed with the result.
Script output: both strategies, dates, interest totals, the avalanche savings]
Which strategy and why, in two sentences — citing the priced difference and the track record, not doctrine]
Debt by debt with projected month — the countdown that makes progress visible]
Bad month: minimums-only protocol, restart trigger] · Windfalls: target-debt rule] · Cliff watch: any promotional deadlines, calendared]
Assumes fixed rates and no new charges — the plan's silent partner is the spending that stops. Educational model, not financial advice — verify with a licensed professional before acting on it.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | pass→fail | 16,921 | 6,268 | -63% | 1 | 1 | 0% | 3,500 | 1,746 | -50% | 0 | 0 | — |
case-02 | fail→pass | 16,535 | 20,871 | +26% | 1 | 1 | 0% | 3,709 | 5,879 | +59% | 0 | 0 | — |
case-03 | fail→fail | 16,337 | 25,561 | +56% | 1 | 1 | 0% | 3,620 | 7,453 | +106% | 0 | 0 | — |
case-15 | pass→pass | 11,889 | 23,427 | +97% | 1 | 1 | 0% | 2,279 | 3,588 | +57% | 0 | 0 | — |
case-04 | pass→pass | 14,383 | 11,229 | -22% | 1 | 1 | 0% | 2,848 | 2,834 | -0% | 0 | 0 | — |
case-05 | pass→fail | 15,058 | 8,366 | -44% | 1 | 1 | 0% | 2,865 | 1,676 | -42% | 0 | 0 | — |
case-06 | pass→pass | 15,277 | 17,816 | +17% | 1 | 1 | 0% | 2,754 | 4,525 | +64% | 0 | 0 | — |
case-07 | fail→pass | 12,629 | 21,570 | +71% | 1 | 1 | 0% | 2,383 | 5,015 | +110% | 0 | 0 | — |
case-08 | fail→fail | 11,327 | 20,987 | +85% | 1 | 1 | 0% | 2,199 | 4,486 | +104% | 0 | 0 | — |
case-09 | pass→pass | 9,482 | 14,791 | +56% | 1 | 1 | 0% | 1,804 | 4,309 | +139% | 0 | 0 | — |
case-10 | fail→pass | 15,891 | 13,430 | -15% | 1 | 1 | 0% | 2,837 | 3,469 | +22% | 0 | 0 | — |
case-11 | pass→pass | 11,171 | 15,698 | +41% | 1 | 1 | 0% | 2,021 | 4,228 | +109% | 0 | 0 | — |
case-12 | fail→fail | 12,422 | 23,489 | +89% | 1 | 1 | 0% | 2,578 | 5,683 | +120% | 0 | 0 | — |
case-13 | pass→pass | 6,734 | 17,230 | +156% | 1 | 1 | 0% | 1,512 | 5,099 | +237% | 0 | 0 | — |
case-14 | pass→pass | 10,779 | 14,606 | +36% | 1 | 1 | 0% | 2,792 | 4,640 | +66% | 0 | 0 | — |
case-16 | pass→pass | 7,660 | 8,645 | +13% | 1 | 1 | 0% | 1,844 | 3,128 | +70% | 0 | 0 | — |
case-17 | fail→fail | 11,821 | 17,255 | +46% | 1 | 1 | 0% | 2,309 | 4,740 | +105% | 0 | 0 | — |
case-18 | pass→pass | 7,861 | 23,870 | +204% | 1 | 1 | 0% | 1,795 | 7,090 | +295% | 0 | 0 | — |
case-19 | pass→pass | 4,987 | 28,840 | +478% | 1 | 1 | 0% | 1,210 | 6,918 | +472% | 0 | 0 | — |
case-20 | pass→pass | 11,945 | 14,452 | +21% | 1 | 1 | 0% | 2,234 | 3,789 | +70% | 0 | 0 | — |
case-21 | fail→fail | 4,374 | 6,172 | +41% | 1 | 1 | 0% | 978 | 1,565 | +60% | 0 | 0 | — |
case-22 | fail→pass | 12,854 | 17,007 | +32% | 1 | 1 | 0% | 2,519 | 4,524 | +80% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 22 cases were attempted, and 19 counted toward the lift figure. The other 3 produced results that are not comparable between the two arms, so they are excluded from the headline rather than averaged into it. The headline lift of +9 percentage points is the difference between those two pass rates over the 19 comparable cases. 2 cases got worse with the skill loaded, and they are included in that figure.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.