Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Use when auditing slow page loads, heavy assets, or rendering delays related to Optimize Google Tag Manager implementation. Verify the actual bottleneck in DevTools, Lighthouse, or field data before recommending changes.
.claude/skills/thedaviddias-gtm-present/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-10 | ✗→✓ | ▲ Improved | 6% | 0% |
| case-20 | ✓→✗ | ▼ Worse | -7% | 0% |
| case-02 | ✓→✓ | ＝ Same ✓ | 20% | 0% |
| case-04 | ✓→✓ | ＝ Same ✓ | -10% | 0% |
| case-05 | ✓→✓ | ＝ Same ✓ | -24% | 0% |
An unoptimized GTM setup can lead to significant main-thread blocking and increased page weight, slowing down the overall user experience.
<head> but audit its impactVerify that GTM is implemented correctly and evaluate its impact on page performance using Lighthouse.
Clean up unused tags, use asynchronous loading, and consider moving to a server-side implementation.
Explain how GTM affects performance and how to balance tracking needs with speed.
Review the routes, assets, and loading behavior that affect Optimize Google Tag Manager implementation. Flag exact files, requests, or rendering steps that add unnecessary network, CPU, or layout cost, and describe the measurement method used to confirm the issue.
For full implementation details, code examples, and framework-specific guidance, see references/rule.md.
Rule page: https://frontendchecklist.io/en/rules/performance/gtm-present
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 13,593 | 3,549 | -74% | 1 | 1 | 0% | 2,750 | 634 | -77% | 0 | 0 | — |
case-02 | pass→pass | 6,271 | 6,893 | +10% | 1 | 1 | 0% | 1,374 | 1,647 | +20% | 0 | 0 | — |
case-03 | fail→fail | 12,966 | 13,375 | +3% | 1 | 1 | 0% | 2,467 | 2,738 | +11% | 0 | 0 | — |
case-04 | pass→pass | 8,990 | 6,333 | -30% | 1 | 1 | 0% | 1,729 | 1,550 | -10% | 0 | 0 | — |
case-05 | pass→pass | 14,368 | 9,717 | -32% | 1 | 1 | 0% | 2,607 | 1,986 | -24% | 0 | 0 | — |
case-15 | pass→pass | 16,381 | 20,917 | +28% | 1 | 1 | 0% | 2,733 | 3,820 | +40% | 0 | 0 | — |
case-06 | pass→pass | 9,678 | 5,034 | -48% | 1 | 1 | 0% | 1,780 | 1,092 | -39% | 0 | 0 | — |
case-07 | pass→pass | 11,619 | 16,676 | +44% | 1 | 1 | 0% | 2,176 | 3,532 | +62% | 0 | 0 | — |
case-08 | pass→pass | 12,350 | 14,751 | +19% | 1 | 1 | 0% | 2,382 | 3,134 | +32% | 0 | 0 | — |
case-09 | pass→pass | 8,294 | 7,203 | -13% | 1 | 1 | 0% | 1,420 | 1,473 | +4% | 0 | 0 | — |
case-10 | fail→pass | 7,224 | 5,996 | -17% | 1 | 1 | 0% | 1,419 | 1,508 | +6% | 0 | 0 | — |
case-11 | pass→pass | 12,777 | 11,091 | -13% | 1 | 1 | 0% | 2,494 | 2,117 | -15% | 0 | 0 | — |
case-12 | pass→pass | 11,990 | 9,641 | -20% | 1 | 1 | 0% | 2,380 | 2,168 | -9% | 0 | 0 | — |
case-13 | pass→pass | 12,822 | 11,741 | -8% | 1 | 1 | 0% | 2,522 | 2,529 | +0% | 0 | 0 | — |
case-14 | pass→pass | 15,990 | 14,531 | -9% | 1 | 1 | 0% | 2,601 | 2,669 | +3% | 0 | 0 | — |
case-16 | pass→pass | 15,968 | 14,563 | -9% | 1 | 1 | 0% | 2,720 | 2,703 | -1% | 0 | 0 | — |
case-17 | pass→pass | 13,745 | 13,998 | +2% | 1 | 1 | 0% | 2,385 | 2,634 | +10% | 0 | 0 | — |
case-18 | pass→pass | 12,463 | 16,424 | +32% | 1 | 1 | 0% | 2,314 | 2,404 | +4% | 0 | 0 | — |
case-19 | fail→fail | 13,746 | 12,465 | -9% | 1 | 1 | 0% | 2,507 | 2,476 | -1% | 0 | 0 | — |
case-20 | pass→fail | 10,469 | 7,204 | -31% | 1 | 1 | 0% | 2,085 | 1,932 | -7% | 0 | 0 | — |
case-21 | pass→pass | 11,094 | 9,121 | -18% | 1 | 1 | 0% | 2,232 | 2,084 | -7% | 0 | 0 | — |
case-22 | pass→pass | 8,138 | 7,931 | -3% | 1 | 1 | 0% | 1,659 | 1,835 | +11% | 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 0 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.