Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Prepare YOUR data room for a fundraise, acquisition, or bank diligence -- builds the checklist for your deal stage, sweeps your folders for what you already have, produces a gap report, and organizes everything into the structure buyers and investors expect. The sell-side complement to cowork-deal-room.
.claude/skills/onewave-ai-cowork-data-room-builder/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 831% | 0% |
| case-02 | ✗→✓ | ▲ Improved | 1955% | 0% |
| case-04 | ✗→✓ | ▲ Improved | 6% | 0% |
| case-06 | ✗→✓ | ▲ Improved | -7% | 0% |
| case-07 | ✗→✓ | ▲ Improved | -7% | 0% |
Prepare a data room the way a banker's analyst does: know what the other side's diligence team will ask for before they ask, find it or flag it, and present it in the structure they expect. Inputs: the deal type (seed/Series A/growth round, acquisition, loan) and one or more folders of company documents.
FINAL (executed/signed), DRAFT, or STALE (superseded or older than the period requested). One file can satisfy multiple items.HAVE, HAVE-BUT-DRAFT, HAVE-BUT-STALE, MISSING. Rank the missing items by how early diligence will hit them (corporate formation docs and cap table before customer contracts). This report is the deliverable that saves the deal timeline.3.2-msa-acme-corp-executed-2025.pdf), and write INDEX.md mapping every checklist item to its file.DRAFT in the filename and index.| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 6,344 | 32,535 | +413% | 1 | 1 | 0% | 437 | 4,069 | +831% | 0 | 0 | — |
case-02 | fail→pass | 3,358 | 36,736 | +994% | 1 | 1 | 0% | 317 | 6,513 | +1955% | 0 | 0 | — |
case-03 | fail→fail | 2,995 | 17,489 | +484% | 1 | 1 | 0% | 409 | 2,998 | +633% | 0 | 0 | — |
case-04 | fail→pass | 15,266 | 13,117 | -14% | 1 | 1 | 0% | 2,290 | 2,416 | +6% | 0 | 0 | — |
case-05 | pass→pass | 11,440 | 5,759 | -50% | 1 | 1 | 0% | 1,548 | 1,490 | -4% | 0 | 0 | — |
case-06 | fail→pass | 10,894 | 8,084 | -26% | 1 | 1 | 0% | 1,887 | 1,752 | -7% | 0 | 0 | — |
case-07 | fail→pass | 11,896 | 7,595 | -36% | 1 | 1 | 0% | 1,858 | 1,734 | -7% | 0 | 0 | — |
case-08 | fail→pass | 8,393 | 8,298 | -1% | 1 | 1 | 0% | 1,619 | 1,850 | +14% | 0 | 0 | — |
case-09 | pass→fail | 12,551 | 11,011 | -12% | 1 | 1 | 0% | 1,788 | 2,242 | +25% | 0 | 0 | — |
case-10 | fail→pass | 9,022 | 5,741 | -36% | 1 | 1 | 0% | 1,431 | 1,521 | +6% | 0 | 0 | — |
case-11 | fail→pass | 9,449 | 8,792 | -7% | 1 | 1 | 0% | 1,411 | 1,870 | +33% | 0 | 0 | — |
case-12 | pass→pass | 11,870 | 8,957 | -25% | 1 | 1 | 0% | 1,791 | 1,938 | +8% | 0 | 0 | — |
case-13 | pass→pass | 9,143 | 6,266 | -31% | 1 | 1 | 0% | 1,348 | 1,612 | +20% | 0 | 0 | — |
case-14 | pass→pass | 11,586 | 7,034 | -39% | 1 | 1 | 0% | 1,681 | 1,687 | +0% | 0 | 0 | — |
case-15 | pass→pass | 11,587 | 6,332 | -45% | 1 | 1 | 0% | 1,930 | 1,658 | -14% | 0 | 0 | — |
case-16 | pass→pass | 15,403 | 11,252 | -27% | 1 | 1 | 0% | 2,480 | 2,427 | -2% | 0 | 0 | — |
case-17 | pass→pass | 10,678 | 10,959 | +3% | 1 | 1 | 0% | 1,584 | 2,265 | +43% | 0 | 0 | — |
case-18 | pass→pass | 12,301 | 12,427 | +1% | 1 | 1 | 0% | 1,752 | 2,403 | +37% | 0 | 0 | — |
case-19 | fail→fail | 24,259 | 20,545 | -15% | 1 | 1 | 0% | 4,009 | 4,181 | +4% | 0 | 0 | — |
case-20 | fail→fail | 5,841 | 11,868 | +103% | 1 | 1 | 0% | 281 | 2,639 | +839% | 0 | 0 | — |
case-21 | pass→pass | 16,548 | 7,669 | -54% | 1 | 1 | 0% | 2,585 | 1,799 | -30% | 0 | 0 | — |
case-22 | pass→pass | 5,500 | 8,625 | +57% | 1 | 1 | 0% | 871 | 1,868 | +114% | 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 21 counted toward the lift figure. The other 1 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 +32 percentage points is the difference between those two pass rates over the 21 comparable cases. 1 case got worse with the skill loaded, and it is 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.