Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Use when evaluating content pages for SEO quality. Applies to blog posts, product descriptions, landing pages, or any page expected to receive organic search traffic.
.claude/skills/thedaviddias-quality/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-11 | ✗→✓ | ▲ Improved | 21% | 0% |
| case-02 | ✗→✓ | ▲ Improved | 19% | 0% |
| case-03 | ✗→✓ | ▲ Improved | 4% | 0% |
| case-15 | ✗→✓ | ▲ Improved | 14% | 0% |
| case-19 | ✗→✓ | ▲ Improved | 25% | 0% |
Google's Search Quality Rater Guidelines and Helpful Content system reward pages that demonstrate genuine expertise and fully satisfy user intent—low-quality pages are algorithmically suppressed site-wide.
Evaluate this page's content for E-E-A-T signals: Does it demonstrate first-hand experience or expertise? Does it fully answer the apparent search intent? Is the information accurate, specific, and supported by evidence? Is there meaningful depth beyond what competitors cover? Identify specific quality gaps.
Identify the primary search intent for this page. Rewrite or expand the content to: (1) fully satisfy that intent, (2) add specific details, examples, or data not found on competing pages, (3) attribute claims to authoritative sources, (4) add author byline with relevant credentials, (5) update any outdated information.
Google's ranking systems are designed to surface content that demonstrates Experience, Expertise, Authoritativeness, and Trustworthiness (E-E-A-T). Content that is generic, lacks depth, or does not fully answer user questions is classified as 'unhelpful' and deprioritised in rankings. High-quality content is the most durable long-term SEO investment.
Analyse the page content for quality signals: (1) Word count — is the topic covered in appropriate depth? (2) Specificity — are claims backed by data or examples? (3) Authorship — is a named author with credentials visible? (4) Freshness — does the page show a publication or updated date? (5) Sources — are external authoritative sources cited? (6) Originality — does the content offer unique insights not found on competing pages?
For full implementation details, code examples, and framework-specific guidance, see references/rule.md.
Rule page: https://frontendchecklist.io/en/rules/seo/quality
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-11 | fail→pass | 16,081 | 15,411 | -4% | 1 | 1 | 0% | 2,601 | 3,137 | +21% | 0 | 0 | — |
case-01 | fail→fail | 13,828 | 12,061 | -13% | 1 | 1 | 0% | 2,290 | 2,748 | +20% | 0 | 0 | — |
case-02 | fail→pass | 8,101 | 7,688 | -5% | 1 | 1 | 0% | 1,463 | 1,747 | +19% | 0 | 0 | — |
case-03 | fail→pass | 9,850 | 7,089 | -28% | 1 | 1 | 0% | 1,689 | 1,760 | +4% | 0 | 0 | — |
case-04 | pass→pass | 8,709 | 6,790 | -22% | 1 | 1 | 0% | 1,495 | 1,646 | +10% | 0 | 0 | — |
case-05 | fail→fail | 5,652 | 5,012 | -11% | 1 | 1 | 0% | 946 | 1,404 | +48% | 0 | 0 | — |
case-06 | pass→pass | 10,859 | 12,663 | +17% | 1 | 1 | 0% | 1,742 | 2,576 | +48% | 0 | 0 | — |
case-07 | pass→fail | 8,133 | 6,567 | -19% | 1 | 1 | 0% | 1,508 | 1,691 | +12% | 0 | 0 | — |
case-08 | pass→pass | 10,335 | 12,426 | +20% | 1 | 1 | 0% | 1,973 | 2,780 | +41% | 0 | 0 | — |
case-09 | pass→pass | 2,929 | 2,001 | -32% | 1 | 1 | 0% | 576 | 854 | +48% | 0 | 0 | — |
case-10 | pass→pass | 9,631 | 6,323 | -34% | 1 | 1 | 0% | 1,622 | 1,542 | -5% | 0 | 0 | — |
case-12 | pass→pass | 12,670 | 16,533 | +30% | 1 | 1 | 0% | 2,212 | 3,320 | +50% | 0 | 0 | — |
case-13 | pass→pass | 2,291 | 2,959 | +29% | 1 | 1 | 0% | 388 | 1,011 | +161% | 0 | 0 | — |
case-14 | pass→pass | 11,402 | 11,154 | -2% | 1 | 1 | 0% | 1,910 | 2,365 | +24% | 0 | 0 | — |
case-15 | fail→pass | 8,276 | 6,366 | -23% | 1 | 1 | 0% | 1,460 | 1,667 | +14% | 0 | 0 | — |
case-16 | fail→fail | 11,826 | 9,528 | -19% | 1 | 1 | 0% | 2,005 | 2,120 | +6% | 0 | 0 | — |
case-17 | pass→pass | 12,917 | 10,396 | -20% | 1 | 1 | 0% | 2,122 | 2,444 | +15% | 0 | 0 | — |
case-18 | fail→fail | 12,490 | 11,244 | -10% | 1 | 1 | 0% | 1,899 | 2,331 | +23% | 0 | 0 | — |
case-19 | fail→pass | 15,639 | 15,425 | -1% | 1 | 1 | 0% | 2,433 | 3,032 | +25% | 0 | 0 | — |
case-20 | pass→pass | 7,308 | 6,550 | -10% | 1 | 1 | 0% | 1,370 | 1,687 | +23% | 0 | 0 | — |
case-21 | pass→pass | 2,415 | 2,295 | -5% | 1 | 1 | 0% | 428 | 900 | +110% | 0 | 0 | — |
case-22 | pass→pass | 2,747 | 2,168 | -21% | 1 | 1 | 0% | 478 | 835 | +75% | 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 +18 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.