Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Build an insurance-grade home inventory from a folder of photos and receipts -- identifies items, pulls values from receipts, estimates replacement costs, organizes by room, and outputs the documentation an insurance claim actually requires. Update mode keeps it current after new purchases.
.claude/skills/onewave-ai-cowork-home-inventory/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-06 | ✗→✓ | ▲ Improved | 11% | 0% |
| case-10 | ✗→✓ | ▲ Improved | 5% | 0% |
| case-14 | ✗→✓ | ▲ Improved | 1% | 0% |
| case-15 | ✗→✓ | ▲ Improved | 17% | 0% |
| case-20 | ✗→✓ | ▲ Improved | -12% | 0% |
Build the document everyone wishes they had after the fire, flood, or break-in: a room-by-room inventory with values and proof. Input: a folder of photos (room shots, close-ups, serial-number shots) and receipts (PDF, email exports, images). Output: a structured inventory file plus a gap list of what still needs documenting.
ESTIMATE. Never present an estimate as a documented value.home-inventory.csv (or .xlsx via the xlsx skill) -- room, item, brand/model, serial, purchase date, purchase price, replacement estimate, documentation status, photo filename(s), receipt filename -- plus home-inventory-summary.md: totals by room and category, the high-value items list, and overall documented vs. estimated ratio.Re-run on the same folder after adding new photos/receipts: append new items, update matched ones, preserve manual edits to the CSV, and report what changed. As a quarterly Cowork scheduled task, sweep the receipts/email-export folder and deliver the diff.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 11,839 | 6,322 | -47% | 1 | 1 | 0% | 2,104 | 1,462 | -31% | 0 | 0 | — |
case-02 | fail→fail | 33,762 | 5,797 | -83% | 1 | 1 | 0% | 4,323 | 1,048 | -76% | 0 | 0 | — |
case-03 | pass→pass | 8,828 | 10,561 | +20% | 1 | 1 | 0% | 1,313 | 2,125 | +62% | 0 | 0 | — |
case-04 | pass→pass | 9,351 | 10,677 | +14% | 1 | 1 | 0% | 1,298 | 2,184 | +68% | 0 | 0 | — |
case-05 | pass→pass | 9,094 | 12,547 | +38% | 1 | 1 | 0% | 1,378 | 2,368 | +72% | 0 | 0 | — |
case-06 | fail→pass | 10,516 | 7,361 | -30% | 1 | 1 | 0% | 1,699 | 1,878 | +11% | 0 | 0 | — |
case-07 | pass→pass | 11,640 | 8,244 | -29% | 1 | 1 | 0% | 1,779 | 2,014 | +13% | 0 | 0 | — |
case-08 | fail→fail | 21,923 | 5,971 | -73% | 1 | 1 | 0% | 1,935 | 1,603 | -17% | 0 | 0 | — |
case-09 | pass→pass | 10,889 | 9,433 | -13% | 1 | 1 | 0% | 1,602 | 2,120 | +32% | 0 | 0 | — |
case-10 | fail→pass | 12,615 | 8,488 | -33% | 1 | 1 | 0% | 2,019 | 2,128 | +5% | 0 | 0 | — |
case-11 | pass→pass | 8,844 | 7,293 | -18% | 1 | 1 | 0% | 1,346 | 1,726 | +28% | 0 | 0 | — |
case-12 | fail→fail | 11,063 | 6,826 | -38% | 1 | 1 | 0% | 1,740 | 1,759 | +1% | 0 | 0 | — |
case-13 | pass→pass | 12,676 | 9,630 | -24% | 1 | 1 | 0% | 2,146 | 2,236 | +4% | 0 | 0 | — |
case-14 | fail→pass | 12,314 | 8,941 | -27% | 1 | 1 | 0% | 2,087 | 2,113 | +1% | 0 | 0 | — |
case-15 | fail→pass | 13,060 | 10,894 | -17% | 1 | 1 | 0% | 1,968 | 2,307 | +17% | 0 | 0 | — |
case-16 | pass→pass | 12,255 | 10,703 | -13% | 1 | 1 | 0% | 2,005 | 2,153 | +7% | 0 | 0 | — |
case-17 | pass→fail | 15,715 | 4,158 | -74% | 1 | 1 | 0% | 2,353 | 1,406 | -40% | 0 | 0 | — |
case-18 | pass→pass | 9,675 | 5,455 | -44% | 1 | 1 | 0% | 1,452 | 1,496 | +3% | 0 | 0 | — |
case-19 | pass→pass | 10,619 | 7,785 | -27% | 1 | 1 | 0% | 1,597 | 1,653 | +4% | 0 | 0 | — |
case-20 | fail→pass | 12,634 | 6,472 | -49% | 1 | 1 | 0% | 1,955 | 1,723 | -12% | 0 | 0 | — |
case-21 | fail→fail | 11,756 | 6,980 | -41% | 1 | 1 | 0% | 1,816 | 1,766 | -3% | 0 | 0 | — |
case-22 | pass→pass | 8,587 | 8,453 | -2% | 1 | 1 | 0% | 1,191 | 1,915 | +61% | 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 +18 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.