Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Compute who gets what at each exit price from a cap table — liquidation preferences, conversion points, and where the founders' share collapses. Use when asked to model an exit waterfall, what do I get if we sell for X, explain liquidation preferences on my cap table, or compare payouts across exit prices. Produces a per-stakeholder payout table across exit values with conversion decisions shown, plus the plain-English reading of what the structure means for each party.
.claude/skills/mohitagw15856-exit-waterfall/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 35% | 0% |
| case-03 | ✗→✓ | ▲ Improved | 25% | 0% |
| case-20 | ✗→✓ | ▲ Improved | -3% | 0% |
| case-22 | ✓→✓ | = Same ✓ | -6% | 0% |
| case-04 | ✓→✓ | = Same ✓ | 47% | 0% |
A cap table is a promise about percentages; an exit waterfall is what actually happens to the money. The difference — liquidation preferences, participation, option strikes — routinely stuns founders at the worst possible moment. This skill computes the waterfall across exit prices and, more importantly, translates it: the price below which your shares are worth nothing, and the price where everyone finally converts.
Ask for these if not provided:
The math ships as a deterministic script — run it rather than arithmetic-by-hand:
bashpython3 scripts/exit_waterfall.py cap.json python3 scripts/exit_waterfall.py cap.json --exit 50000000 --json
Input shape and worked example are in the script docstring. It brute-forces the conversion equilibrium (each non-participating preferred converts only when as-converted beats its preference) and applies the treasury method to options. Stated simplifications: preferences are pari passu (no seniority stacking) and participation is uncapped — flag both when the real cap table differs, and say the numbers shift accordingly.
The script's table: exit price × stakeholder, with the converts column]
Pari passu, uncapped participation, valuation of the classes as supplied — every deviation from the real documents named.]
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-22 | pass→pass | 52,955 | 15,925 | -70% | 1 | 1 | 0% | 3,692 | 3,464 | -6% | 0 | 0 | — |
case-01 | fail→pass | 22,421 | 25,244 | +13% | 1 | 1 | 0% | 4,598 | 6,189 | +35% | 0 | 0 | — |
case-02 | fail→fail | 30,210 | 29,424 | -3% | 1 | 1 | 0% | 6,303 | 7,125 | +13% | 0 | 0 | — |
case-03 | fail→pass | 21,857 | 23,063 | +6% | 1 | 1 | 0% | 4,454 | 5,577 | +25% | 0 | 0 | — |
case-04 | pass→pass | 10,870 | 10,080 | -7% | 1 | 1 | 0% | 1,662 | 2,449 | +47% | 0 | 0 | — |
case-05 | pass→pass | 9,909 | 9,162 | -8% | 1 | 1 | 0% | 1,450 | 2,183 | +51% | 0 | 0 | — |
case-06 | pass→pass | 8,828 | 10,486 | +19% | 1 | 1 | 0% | 1,812 | 2,925 | +61% | 0 | 0 | — |
case-07 | pass→pass | 7,458 | 9,596 | +29% | 1 | 1 | 0% | 1,531 | 2,748 | +79% | 0 | 0 | — |
case-08 | pass→pass | 16,181 | 23,231 | +44% | 1 | 1 | 0% | 3,112 | 5,586 | +79% | 0 | 0 | — |
case-09 | pass→pass | 8,796 | 18,496 | +110% | 1 | 1 | 0% | 1,707 | 4,623 | +171% | 0 | 0 | — |
case-10 | pass→pass | 6,426 | 11,802 | +84% | 1 | 1 | 0% | 1,056 | 2,995 | +184% | 0 | 0 | — |
case-11 | pass→pass | 13,522 | 7,041 | -48% | 1 | 1 | 0% | 2,304 | 2,035 | -12% | 0 | 0 | — |
case-12 | pass→pass | 8,110 | 2,123 | -74% | 1 | 1 | 0% | 1,211 | 1,135 | -6% | 0 | 0 | — |
case-13 | pass→pass | 5,588 | 3,447 | -38% | 1 | 1 | 0% | 923 | 1,469 | +59% | 0 | 0 | — |
case-14 | pass→pass | 12,274 | 10,340 | -16% | 1 | 1 | 0% | 2,095 | 2,608 | +24% | 0 | 0 | — |
case-15 | fail→fail | 11,838 | 23,128 | +95% | 1 | 1 | 0% | 2,541 | 5,638 | +122% | 0 | 0 | — |
case-16 | pass→pass | 13,110 | 9,572 | -27% | 1 | 1 | 0% | 2,301 | 2,559 | +11% | 0 | 0 | — |
case-17 | pass→pass | 8,791 | 10,875 | +24% | 1 | 1 | 0% | 1,913 | 2,950 | +54% | 0 | 0 | — |
case-18 | pass→pass | 13,792 | 12,916 | -6% | 1 | 1 | 0% | 2,407 | 3,169 | +32% | 0 | 0 | — |
case-19 | pass→pass | 8,415 | 5,315 | -37% | 1 | 1 | 0% | 1,316 | 1,668 | +27% | 0 | 0 | — |
case-20 | fail→pass | 13,072 | 12,910 | -1% | 1 | 1 | 0% | 2,531 | 2,456 | -3% | 0 | 0 | — |
case-21 | pass→pass | 13,218 | 7,751 | -41% | 1 | 1 | 0% | 2,159 | 2,121 | -2% | 0 | 0 | — |
case-23 | fail→fail | 22,784 | 31,466 | +38% | 1 | 1 | 0% | 4,603 | 7,031 | +53% | 0 | 0 | — |
case-24 | fail→fail | 17,597 | 18,669 | +6% | 1 | 1 | 0% | 3,378 | 4,231 | +25% | 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. 24 cases were attempted. The headline lift of +13 percentage points is the difference between those two pass rates over the 24 comparable cases.
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.