Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Produce a read-only, evidence-based paid-media performance review across connected platforms or supplied exports. Use for weekly or monthly reports, scorecards, performance, CPA, ROAS, CTR, spend trends, pacing, or conversion-tracking health.
.claude/skills/nowork-studio-paid-ads-review/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-03 | ✗→✓ | ▲ Improved | 165% | 0% |
| case-07 | ✗→✓ | ▲ Improved | 88% | 0% |
| case-09 | ✗→✓ | ▲ Improved | 9% | 0% |
| case-10 | ✗→✓ | ▲ Improved | -15% | 0% |
| case-12 | ✗→✓ | ▲ Improved | 52% | 0% |
Read ../shared/operating-contract.md and ../shared/measurement-framework.md. This is read-only.
Check which sources are actually connected, request the missing export rather than reporting a failed call, and use the most recent complete equivalent window. For every source, keep currency, conversion definition, attribution window, and reporting lag visible. Verify tracking before treating CPA, ROAS, or revenue as a decision-grade metric.
For live platform accounts, hand data collection and diagnostics to /notfair:google-ads, /notfair:meta-ads, /notfair:paid-ads-x, /notfair:paid-ads-linkedin, /notfair:paid-ads-reddit, or /notfair:paid-ads-tiktok. Keep Amazon and ChatGPT Ads review grounded in a verified connector or supplied platform export.
Lead with the strongest contributor, the largest risk, and one recommended next action. Include a platform scorecard with spend, qualified conversions, CPA, attributable revenue/ROAS where available, link CTR, and pace against the declared budget. Compare each row to the preceding equivalent period and name the likely driver only when data supports it.
Do not present an unqualified blended CPA or ROAS. If a cross-channel aggregate is useful, label the consistent conversion definition, attribution source, spend-weighted formula, and included channels. Separate confirmed facts from inference and mark absent tracking or data as a blocking limitation.
End with hold, investigate, or one proposed action; route mutations to /notfair:paid-ads-optimize or the relevant platform operator skill.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 15,112 | 16,536 | +9% | 1 | 1 | 0% | 2,449 | 3,065 | +25% | 0 | 0 | — |
case-02 | fail→fail | 10,704 | 15,978 | +49% | 1 | 1 | 0% | 1,526 | 2,332 | +53% | 0 | 0 | — |
case-03 | fail→pass | 10,029 | 22,295 | +122% | 1 | 1 | 0% | 1,574 | 4,178 | +165% | 0 | 0 | — |
case-04 | pass→pass | 26,580 | 19,136 | -28% | 1 | 1 | 0% | 1,708 | 1,757 | +3% | 0 | 0 | — |
case-05 | pass→pass | 16,949 | 16,102 | -5% | 1 | 1 | 0% | 2,679 | 2,852 | +6% | 0 | 0 | — |
case-06 | pass→pass | 14,205 | 23,330 | +64% | 1 | 1 | 0% | 2,203 | 2,694 | +22% | 0 | 0 | — |
case-07 | fail→pass | 8,820 | 12,236 | +39% | 1 | 1 | 0% | 1,327 | 2,492 | +88% | 0 | 0 | — |
case-08 | pass→pass | 20,912 | 10,873 | -48% | 1 | 1 | 0% | 1,742 | 1,978 | +14% | 0 | 0 | — |
case-09 | fail→pass | 17,439 | 18,577 | +7% | 1 | 1 | 0% | 2,816 | 3,080 | +9% | 0 | 0 | — |
case-10 | fail→pass | 10,919 | 7,024 | -36% | 1 | 1 | 0% | 1,613 | 1,375 | -15% | 0 | 0 | — |
case-11 | pass→pass | 9,380 | 14,489 | +54% | 1 | 1 | 0% | 1,430 | 2,436 | +70% | 0 | 0 | — |
case-12 | fail→pass | 7,050 | 8,254 | +17% | 1 | 1 | 0% | 1,000 | 1,524 | +52% | 0 | 0 | — |
case-13 | fail→fail | 9,547 | 9,257 | -3% | 1 | 1 | 0% | 1,137 | 1,761 | +55% | 0 | 0 | — |
case-14 | fail→pass | 126,398 | 14,442 | -89% | 1 | 1 | 0% | 768 | 2,468 | +221% | 0 | 0 | — |
case-15 | fail→fail | 2,780 | 8,110 | +192% | 1 | 1 | 0% | 286 | 1,365 | +377% | 0 | 0 | — |
case-16 | pass→pass | 29,183 | 13,118 | -55% | 1 | 1 | 0% | 2,369 | 2,323 | -2% | 0 | 0 | — |
case-17 | fail→fail | 12,605 | 23,689 | +88% | 1 | 1 | 0% | 1,890 | 2,132 | +13% | 0 | 0 | — |
case-18 | fail→pass | 9,959 | 18,979 | +91% | 1 | 1 | 0% | 1,447 | 3,323 | +130% | 0 | 0 | — |
case-19 | fail→pass | 18,526 | 18,550 | +0% | 1 | 1 | 0% | 2,531 | 2,865 | +13% | 0 | 0 | — |
case-20 | fail→pass | 59,352 | 20,055 | -66% | 1 | 1 | 0% | 1,867 | 3,393 | +82% | 0 | 0 | — |
case-21 | fail→pass | 10,960 | 31,373 | +186% | 1 | 1 | 0% | 1,639 | 3,047 | +86% | 0 | 0 | — |
case-22 | fail→fail | 7,636 | 25,611 | +235% | 1 | 1 | 0% | 1,021 | 1,788 | +75% | 0 | 0 | — |
case-23 | fail→fail | 21,038 | 18,976 | -10% | 1 | 1 | 0% | 3,093 | 3,026 | -2% | 0 | 0 | — |
case-24 | fail→pass | 8,344 | 13,178 | +58% | 1 | 1 | 0% | 1,203 | 2,469 | +105% | 0 | 0 | — |
case-25 | fail→pass | 11,217 | 16,355 | +46% | 1 | 1 | 0% | 1,688 | 2,865 | +70% | 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. 25 cases were attempted. The headline lift of +48 percentage points is the difference between those two pass rates over the 25 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.
| Model | Method | Date | Lift |
|---|---|---|---|
| gemini-3.6-flash | verified | 8/29/2026 | 0% |
Other measured skills in the registry, with their headline benchmark lift.