Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →TikTok Ads deep analysis covering creative quality, tracking, bidding, campaign structure, and TikTok Shop. Evaluates 25 checks with emphasis on creative-first strategy, safe zone compliance, and Smart+ campaigns. Triggers on: "TikTok Ads", "TikTok marketing", "TikTok Shop", "Spark Ads", "Smart+", "TikTok campaign", "TikTok audit"
.claude/skills/miosa-osa-ads-tiktok/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-02 | ✗→✓ | ▲ Improved | 8% | 0% |
| case-03 | ✗→✓ | ▲ Improved | 34% | 0% |
| case-04 | ✗→✓ | ▲ Improved | 62% | 0% |
| case-05 | ✗→✓ | ▲ Improved | 51% | 0% |
| case-07 | ✗→✓ | ▲ Improved | 73% | 0% |
> 25-check audit of TikTok Ads with creative-first strategy and Shop assessment.
bash/ads tiktok
TikTok success depends primarily on creative quality.
+----------------------+
| UNSAFE (status) | Y: 0-150px
+----------------------+
| |UNSA|
| SAFE ZONE |FE | X: 40-940px
| 900x1320px |icon| Y: 150-1470px
| | |
+----------------------+ Right 140px: like/comment/share
| UNSAFE (caption) | Y: 1470-1920px
+----------------------+For e-commerce:
| Metric | Pass | Warning | Fail | |--------|------|---------|------| | CTR (in-feed) | >=1.0% | 0.5-1.0% | <0.5% | | Creatives per ad group | >=6 | 3-5 | <3 | | Video watch time | >=6s | 3-6s | <3s | | Learning conversions | >=50/week | 30-50/week | <30/week | | Daily budget | >=50x CPA | 20-49x CPA | <20x CPA | | Creative age (declining) | <7 days | 7-14 days | >14 days |
TIKTOK-ADS-REPORT.md — Full 25-check findings with pass/warning/fail| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 52,419 | 40,631 | -22% | 1 | 1 | 0% | 6,369 | 8,477 | +33% | 0 | 0 | — |
case-02 | fail→pass | 58,120 | 39,596 | -32% | 1 | 1 | 0% | 8,258 | 8,924 | +8% | 0 | 0 | — |
case-03 | fail→pass | 51,132 | 39,776 | -22% | 1 | 1 | 0% | 6,171 | 8,252 | +34% | 0 | 0 | — |
case-04 | fail→pass | 13,521 | 12,607 | -7% | 1 | 1 | 0% | 1,924 | 3,111 | +62% | 0 | 0 | — |
case-05 | fail→pass | 29,765 | 13,913 | -53% | 1 | 1 | 0% | 2,056 | 3,103 | +51% | 0 | 0 | — |
case-06 | pass→pass | 17,290 | 11,688 | -32% | 1 | 1 | 0% | 1,641 | 3,037 | +85% | 0 | 0 | — |
case-07 | fail→pass | 21,294 | 11,033 | -48% | 1 | 1 | 0% | 1,784 | 3,084 | +73% | 0 | 0 | — |
case-08 | fail→pass | 15,273 | 8,704 | -43% | 1 | 1 | 0% | 2,047 | 2,637 | +29% | 0 | 0 | — |
case-09 | fail→pass | 14,840 | 17,826 | +20% | 1 | 1 | 0% | 2,093 | 3,646 | +74% | 0 | 0 | — |
case-10 | fail→pass | 27,118 | 15,455 | -43% | 1 | 1 | 0% | 2,329 | 3,458 | +48% | 0 | 0 | — |
case-11 | pass→pass | 12,651 | 8,513 | -33% | 1 | 1 | 0% | 1,668 | 2,603 | +56% | 0 | 0 | — |
case-12 | pass→pass | 11,186 | 11,773 | +5% | 1 | 1 | 0% | 1,671 | 3,102 | +86% | 0 | 0 | — |
case-13 | pass→pass | 14,638 | 14,316 | -2% | 1 | 1 | 0% | 2,172 | 3,411 | +57% | 0 | 0 | — |
case-14 | fail→pass | 13,254 | 6,383 | -52% | 1 | 1 | 0% | 2,053 | 2,423 | +18% | 0 | 0 | — |
case-15 | pass→pass | 12,662 | 16,888 | +33% | 1 | 1 | 0% | 1,996 | 2,954 | +48% | 0 | 0 | — |
case-16 | pass→pass | 11,382 | 57,152 | +402% | 1 | 1 | 0% | 1,792 | 2,130 | +19% | 0 | 0 | — |
case-17 | pass→pass | 13,315 | 9,598 | -28% | 1 | 1 | 0% | 1,853 | 2,708 | +46% | 0 | 0 | — |
case-18 | pass→pass | 17,746 | 10,335 | -42% | 1 | 1 | 0% | 2,286 | 2,923 | +28% | 0 | 0 | — |
case-19 | pass→pass | 12,371 | 7,179 | -42% | 1 | 1 | 0% | 2,003 | 2,287 | +14% | 0 | 0 | — |
case-20 | fail→pass | 12,989 | 10,195 | -22% | 1 | 1 | 0% | 2,112 | 2,583 | +22% | 0 | 0 | — |
case-21 | pass→pass | 20,305 | 28,539 | +41% | 1 | 1 | 0% | 3,379 | 5,383 | +59% | 0 | 0 | — |
case-22 | pass→pass | 13,699 | 28,461 | +108% | 1 | 1 | 0% | 1,896 | 3,485 | +84% | 0 | 0 | — |
case-23 | pass→pass | 17,706 | 19,656 | +11% | 1 | 1 | 0% | 3,566 | 4,705 | +32% | 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. 23 cases were attempted. The headline lift of +43 percentage points is the difference between those two pass rates over the 23 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.