Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Drop a folder of vendor quotes, proposals, and SOWs on Cowork -- normalizes them into one comparison matrix, computes true total cost of ownership, surfaces the terms each vendor buried, and arms you with negotiation leverage points and reference-check questions.
.claude/skills/onewave-ai-cowork-vendor-comparison/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-06 | ✗→✓ | ▲ Improved | 27% | 0% |
| case-16 | ✗→✓ | ▲ Improved | 6% | 0% |
| case-03 | ✓→✓ | ＝ Same ✓ | -1% | 0% |
| case-07 | ✓→✓ | ＝ Same ✓ | -6% | 0% |
| case-08 | ✓→✓ | ＝ Same ✓ | 22% | 0% |
Compare vendor proposals the way a procurement lead does: normalize everything to the same units before judging anything, price the full term not the first invoice, and read the terms the sales deck skipped. Input: a folder of quotes/proposals/SOWs (PDF, .docx, .xlsx) and, ideally, one sentence on what the purchase needs to accomplish.
UNPRICED -- ask, never zero. Unpriced is not free.vendor-comparison.md: side-by-side table of cost, capability fit against the stated need, SLA strength, implementation risk, and contract flexibility -- each cell citing the source document and page. Follow with a plain-English paragraph per vendor: the honest case for and against.| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 8,832 | 39,789 | +351% | 1 | 1 | 0% | 1,436 | 6,045 | +321% | 0 | 0 | — |
case-02 | fail→fail | 30,905 | 35,909 | +16% | 1 | 1 | 0% | 5,487 | 6,296 | +15% | 0 | 0 | — |
case-03 | pass→pass | 11,106 | 7,568 | -32% | 1 | 1 | 0% | 1,881 | 1,868 | -1% | 0 | 0 | — |
case-04 | fail→fail | 12,898 | 8,945 | -31% | 1 | 1 | 0% | 1,777 | 1,997 | +12% | 0 | 0 | — |
case-05 | fail→fail | 8,189 | 1,578 | -81% | 1 | 1 | 0% | 1,277 | 851 | -33% | 0 | 0 | — |
case-06 | fail→pass | 13,601 | 11,879 | -13% | 1 | 1 | 0% | 1,898 | 2,404 | +27% | 0 | 0 | — |
case-07 | pass→pass | 10,967 | 5,522 | -50% | 1 | 1 | 0% | 1,601 | 1,508 | -6% | 0 | 0 | — |
case-08 | pass→pass | 12,307 | 10,771 | -12% | 1 | 1 | 0% | 1,879 | 2,285 | +22% | 0 | 0 | — |
case-09 | pass→pass | 11,021 | 9,968 | -10% | 1 | 1 | 0% | 1,783 | 2,234 | +25% | 0 | 0 | — |
case-19 | fail→fail | 14,589 | 8,903 | -39% | 1 | 1 | 0% | 2,186 | 2,038 | -7% | 0 | 0 | — |
case-10 | fail→fail | 9,290 | 4,847 | -48% | 1 | 1 | 0% | 1,332 | 1,342 | +1% | 0 | 0 | — |
case-11 | fail→fail | 14,913 | 11,470 | -23% | 1 | 1 | 0% | 2,098 | 2,348 | +12% | 0 | 0 | — |
case-12 | pass→pass | 12,153 | 10,459 | -14% | 1 | 1 | 0% | 1,749 | 2,170 | +24% | 0 | 0 | — |
case-13 | fail→fail | 13,824 | 12,126 | -12% | 1 | 1 | 0% | 1,927 | 2,373 | +23% | 0 | 0 | — |
case-14 | pass→pass | 8,190 | 10,360 | +26% | 1 | 1 | 0% | 1,520 | 2,340 | +54% | 0 | 0 | — |
case-15 | pass→pass | 12,634 | 8,447 | -33% | 1 | 1 | 0% | 1,804 | 1,929 | +7% | 0 | 0 | — |
case-16 | fail→pass | 15,787 | 14,340 | -9% | 1 | 1 | 0% | 2,383 | 2,527 | +6% | 0 | 0 | — |
case-17 | fail→fail | 17,588 | 12,538 | -29% | 1 | 1 | 0% | 2,490 | 2,458 | -1% | 0 | 0 | — |
case-18 | fail→fail | 9,660 | 10,115 | +5% | 1 | 1 | 0% | 1,415 | 2,091 | +48% | 0 | 0 | — |
case-20 | pass→pass | 15,578 | 13,925 | -11% | 1 | 1 | 0% | 2,463 | 2,698 | +10% | 0 | 0 | — |
case-21 | pass→pass | 15,505 | 14,593 | -6% | 1 | 1 | 0% | 2,644 | 2,910 | +10% | 0 | 0 | — |
case-22 | pass→pass | 16,187 | 16,624 | +3% | 1 | 1 | 0% | 2,426 | 3,140 | +29% | 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. The headline lift of +9 percentage points is the difference between those two pass rates over the 22 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.