Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Audit and fix HTML metadata including page titles, meta descriptions, canonical URLs, Open Graph tags, Twitter cards, favicons, JSON-LD structured data, and robots directives. Use when adding or reviewing SEO and social metadata.
.claude/skills/fixing-metadata/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-03 | ✗→✓ | ▲ Improved | 94% | 0% |
| case-08 | ✗→✓ | ▲ Improved | 32% | 0% |
| case-04 | ✗→✓ | ▲ Improved | -22% | 0% |
| case-15 | ✗→✓ | ▲ Improved | 11% | 0% |
| case-01 | ✓→✗ | ▼ Worse | -64% | 0% |
Use this skill when you need audit and fix HTML metadata including page titles, meta descriptions, canonical URLs, Open Graph tags, Twitter cards, favicons, JSON-LD structured data, and robots directives. Use when adding SEO metadata, fixing social share previews, reviewing Open Graph tags, setting up canonical...
Reference these guidelines when:
| priority | category | impact | |----------|----------|--------| | 1 | correctness and duplication | critical | | 2 | title and description | high | | 3 | canonical and indexing | high | | 4 | social cards | high | | 5 | icons and manifest | medium | | 6 | structured data | medium | | 7 | locale and alternates | low-medium | | 8 | tool boundaries | critical |
User request:
> Audit and fix HTML metadata including page titles, meta descriptions, canonical URLs, Open Graph tags, Twitter cards, favicons, JSON-LD structured data, and robots directives.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-02 | pass→pass | 9,009 | 9,208 | +2% | 1 | 1 | 0% | 1,681 | 2,938 | +75% | 0 | 0 | — |
case-03 | fail→pass | 9,541 | 16,157 | +69% | 1 | 1 | 0% | 1,933 | 3,755 | +94% | 0 | 0 | — |
case-08 | fail→pass | 8,853 | 6,006 | -32% | 1 | 1 | 0% | 1,577 | 2,079 | +32% | 0 | 0 | — |
case-01 | pass→fail | 22,032 | 5,800 | -74% | 1 | 1 | 0% | 3,751 | 1,362 | -64% | 0 | 0 | — |
case-04 | fail→pass | 14,654 | 8,728 | -40% | 1 | 1 | 0% | 3,429 | 2,690 | -22% | 0 | 0 | — |
case-05 | pass→pass | 7,842 | 5,804 | -26% | 1 | 1 | 0% | 1,488 | 2,179 | +46% | 0 | 0 | — |
case-06 | pass→pass | 9,495 | 5,338 | -44% | 1 | 1 | 0% | 1,953 | 2,004 | +3% | 0 | 0 | — |
case-07 | pass→pass | 8,557 | 3,424 | -60% | 1 | 1 | 0% | 1,646 | 1,706 | +4% | 0 | 0 | — |
case-09 | pass→pass | 9,650 | 6,094 | -37% | 1 | 1 | 0% | 1,948 | 2,181 | +12% | 0 | 0 | — |
case-10 | pass→pass | 11,516 | 6,396 | -44% | 1 | 1 | 0% | 2,302 | 2,236 | -3% | 0 | 0 | — |
case-11 | pass→pass | 14,019 | 7,695 | -45% | 1 | 1 | 0% | 2,349 | 2,482 | +6% | 0 | 0 | — |
case-12 | pass→pass | 7,066 | 3,266 | -54% | 1 | 1 | 0% | 1,220 | 1,632 | +34% | 0 | 0 | — |
case-13 | pass→pass | 6,576 | 3,324 | -49% | 1 | 1 | 0% | 1,195 | 1,641 | +37% | 0 | 0 | — |
case-14 | pass→pass | 5,236 | 3,905 | -25% | 1 | 1 | 0% | 1,002 | 1,758 | +75% | 0 | 0 | — |
case-15 | fail→pass | 11,130 | 4,548 | -59% | 1 | 1 | 0% | 1,864 | 2,064 | +11% | 0 | 0 | — |
case-16 | pass→pass | 8,777 | 5,929 | -32% | 1 | 1 | 0% | 1,653 | 2,144 | +30% | 0 | 0 | — |
case-17 | pass→pass | 7,199 | 3,779 | -48% | 1 | 1 | 0% | 1,292 | 1,671 | +29% | 0 | 0 | — |
case-18 | pass→pass | 12,034 | 9,811 | -18% | 1 | 1 | 0% | 2,160 | 2,378 | +10% | 0 | 0 | — |
case-19 | pass→pass | 13,156 | 5,494 | -58% | 1 | 1 | 0% | 1,910 | 1,910 | 0% | 0 | 0 | — |
case-20 | pass→pass | 12,069 | 4,781 | -60% | 1 | 1 | 0% | 2,095 | 1,849 | -12% | 0 | 0 | — |
case-21 | pass→pass | 10,683 | 9,784 | -8% | 1 | 1 | 0% | 1,797 | 2,702 | +50% | 0 | 0 | — |
case-22 | pass→pass | 8,637 | 3,983 | -54% | 1 | 1 | 0% | 1,411 | 1,628 | +15% | 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, and 21 counted toward the lift figure. The other 1 produced results that are not comparable between the two arms, so they are excluded from the headline rather than averaged into it. The headline lift of +14 percentage points is the difference between those two pass rates over the 21 comparable cases. 1 case got worse with the skill loaded, and it is included in that figure.
The publisher has shipped newer versions since this run, so these numbers describe v2, not the version currently listed.
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 | 7/30/2026 | — |
Other measured skills in the registry, with their headline benchmark lift.