Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →When the user wants to set up, optimize, or manage TikTok Ads. Also use when the user mentions "TikTok Ads," "TikTok for Business," "TikTok Pixel," "Events API," "TikTok Spark Ads," or "TikTok video ads." For organic TikTok, use tiktok-captions.
.claude/skills/kostja94-tiktok-ads/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-10 | ✗→✓ | ▲ Improved | 1% | 0% |
| case-13 | ✗→✓ | ▲ Improved | 13% | 0% |
| case-02 | ✓→✓ | ＝ Same ✓ | 69% | 0% |
| case-03 | ✓→✓ | ＝ Same ✓ | 113% | 0% |
| case-04 | ✓→✓ | ＝ Same ✓ | -3% | 0% |
Guides TikTok Ads setup, creative strategy, and optimization. TikTok excels at younger demographics (18–34) and viral-style video; use when your audience skews young and you have video creative capacity.
When invoking: On first use, if helpful, open with 1–2 sentences on what this skill covers and why it matters, then provide the main output. On subsequent use or when the user asks to skip, go directly to the main output.
| Format | Use | |--------|-----| | In-Feed | Native video; 9–15s or longer; primary format | | TopView | First impression; premium; awareness | | Spark Ads | Boost organic posts; native; creator partnership | | Collection | Product catalog; e-commerce |
| Type | Use | |------|-----| | Interest | Broad; algorithm-driven | | Demographics | Age; gender; location | | Lookalike | Based on converters or engagers | | Retargeting | Website visitors; video viewers |
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 27,211 | 20,583 | -24% | 1 | 1 | 0% | 3,870 | 3,560 | -8% | 0 | 0 | — |
case-02 | pass→pass | 7,654 | 8,709 | +14% | 1 | 1 | 0% | 1,086 | 1,833 | +69% | 0 | 0 | — |
case-03 | pass→pass | 4,988 | 4,832 | -3% | 1 | 1 | 0% | 696 | 1,480 | +113% | 0 | 0 | — |
case-04 | pass→pass | 14,893 | 11,194 | -25% | 1 | 1 | 0% | 2,270 | 2,211 | -3% | 0 | 0 | — |
case-05 | pass→pass | 11,337 | 9,562 | -16% | 1 | 1 | 0% | 1,744 | 2,317 | +33% | 0 | 0 | — |
case-06 | pass→pass | 10,595 | 10,114 | -5% | 1 | 1 | 0% | 1,581 | 2,131 | +35% | 0 | 0 | — |
case-07 | pass→pass | 11,642 | 11,185 | -4% | 1 | 1 | 0% | 1,673 | 2,158 | +29% | 0 | 0 | — |
case-08 | pass→pass | 15,784 | 15,492 | -2% | 1 | 1 | 0% | 2,312 | 2,825 | +22% | 0 | 0 | — |
case-09 | pass→pass | 12,317 | 10,116 | -18% | 1 | 1 | 0% | 1,974 | 1,975 | +0% | 0 | 0 | — |
case-10 | fail→pass | 13,491 | 8,917 | -34% | 1 | 1 | 0% | 2,065 | 2,082 | +1% | 0 | 0 | — |
case-11 | pass→pass | 13,746 | 9,925 | -28% | 1 | 1 | 0% | 2,462 | 2,225 | -10% | 0 | 0 | — |
case-12 | pass→pass | 15,926 | 15,056 | -5% | 1 | 1 | 0% | 2,376 | 2,695 | +13% | 0 | 0 | — |
case-13 | fail→pass | 10,932 | 10,735 | -2% | 1 | 1 | 0% | 1,943 | 2,193 | +13% | 0 | 0 | — |
case-14 | pass→pass | 12,247 | 7,508 | -39% | 1 | 1 | 0% | 1,986 | 1,912 | -4% | 0 | 0 | — |
case-15 | pass→pass | 5,218 | 5,954 | +14% | 1 | 1 | 0% | 884 | 1,808 | +105% | 0 | 0 | — |
case-16 | pass→pass | 6,584 | 6,202 | -6% | 1 | 1 | 0% | 1,221 | 1,747 | +43% | 0 | 0 | — |
case-17 | pass→pass | 5,725 | 5,918 | +3% | 1 | 1 | 0% | 1,047 | 1,578 | +51% | 0 | 0 | — |
case-18 | pass→pass | 5,554 | 6,597 | +19% | 1 | 1 | 0% | 1,023 | 1,802 | +76% | 0 | 0 | — |
case-19 | pass→pass | 4,491 | 6,890 | +53% | 1 | 1 | 0% | 776 | 1,889 | +143% | 0 | 0 | — |
case-20 | pass→pass | 9,391 | 9,046 | -4% | 1 | 1 | 0% | 1,525 | 2,130 | +40% | 0 | 0 | — |
case-21 | pass→pass | 9,726 | 9,722 | -0% | 1 | 1 | 0% | 1,585 | 2,106 | +33% | 0 | 0 | — |
case-22 | fail→fail | 9,282 | 10,933 | +18% | 1 | 1 | 0% | 1,604 | 2,291 | +43% | 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. 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.