Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Audit a backlink profile using Ahrefs MCP data: profile health, anchor text distribution, toxic link identification, lost link reclamation, and gap analysis against competitors. Use this skill when reviewing backlink health, scoping a disavow project, recovering from a manual action, planning link building, or doing M&A due diligence on a property's link equity. Triggers on backlink audit, link profile, toxic links, disavow, anchor text analysis, lost links, link reclamation, referring domains,
.claude/skills/rampstackco-seo-backlink-audit/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 7% | 0% |
| case-02 | ✗→✓ | ▲ Improved | 26% | 0% |
| case-03 | ✗→✓ | ▲ Improved | 24% | 0% |
| case-04 | ✗→✓ | ▲ Improved | 38% | 0% |
| case-19 | ✗→✓ | ▲ Improved | 64% | 0% |
Audit a backlink profile using data pulled from the Ahrefs MCP. Stack-agnostic. Produces a profile health assessment, a toxic link list (if needed), a reclamation list, and a gap analysis against competitors.
seo-offpage)seo-audit-orchestration)seo-keyword-gap-audit)seo-site-health-audit)A profile is healthy when all five dimensions are healthy. Issues in any one can cap rankings or invite penalties.
Volume of links matters less than the quality and relevance of the linking domains.
Pull from Ahrefs:
Healthy signal: a long tail of relevant, established domains. Concentration in any single source is a risk.
How fast links are gained or lost.
Pull from Ahrefs:
Healthy signal: steady positive growth. Sudden spikes suggest paid campaigns or negative SEO. Sudden drops correlate with redirect mistakes, content removal, or the linking sites going down.
The mix of anchor types tells search engines what the property is about and whether the profile looks natural.
Categories to map:
Healthy signal: branded and naked URL anchors dominate. Exact-match keyword anchors are a small minority. Heavy exact-match concentration triggers spam filters.
Not all links are equal.
Categories:
Healthy signal: editorial in-content dofollow links from relevant sites dominate.
What links do peers have that you do not?
Pull from Ahrefs:
Healthy signal: the gap is shrinking over time. A widening gap is a strategic risk.
references/toxic-link-criteria.md.seo-offpage and seo-audit-orchestration.A backlink audit document with:
Length: 8-15 pages depending on profile complexity.
This skill's output depends on data, measurements, or tool results it cannot generate on its own. When a required input, tool, or data source is unavailable or unverifiable, the sanctioned output is the deliverable with the gap stated: what was needed, what was actually obtained or verified, and which parts of the output are affected. Fabricating, estimating, or interpolating a required number to complete the deliverable is never sanctioned. A stated gap is a complete answer.
references/toxic-link-criteria.md - Decision framework for classifying a link as toxic, including the disavow file format and threshold guidance.| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 33,244 | 27,566 | -17% | 1 | 1 | 0% | 6,274 | 6,695 | +7% | 0 | 0 | — |
case-02 | fail→pass | 35,564 | 35,086 | -1% | 1 | 1 | 0% | 6,290 | 7,917 | +26% | 0 | 0 | — |
case-03 | fail→pass | 35,998 | 35,668 | -1% | 1 | 1 | 0% | 6,258 | 7,754 | +24% | 0 | 0 | — |
case-04 | fail→pass | 20,042 | 17,418 | -13% | 1 | 1 | 0% | 3,186 | 4,389 | +38% | 0 | 0 | — |
case-05 | pass→pass | 9,780 | 11,294 | +15% | 1 | 1 | 0% | 1,654 | 3,465 | +109% | 0 | 0 | — |
case-06 | fail→fail | 11,543 | 8,071 | -30% | 1 | 1 | 0% | 1,983 | 3,033 | +53% | 0 | 0 | — |
case-07 | pass→pass | 11,939 | 11,126 | -7% | 1 | 1 | 0% | 1,930 | 3,399 | +76% | 0 | 0 | — |
case-08 | pass→pass | 15,022 | 15,314 | +2% | 1 | 1 | 0% | 2,246 | 4,084 | +82% | 0 | 0 | — |
case-09 | pass→pass | 10,820 | 12,371 | +14% | 1 | 1 | 0% | 1,770 | 3,542 | +100% | 0 | 0 | — |
case-10 | pass→pass | 11,275 | 7,958 | -29% | 1 | 1 | 0% | 1,879 | 2,905 | +55% | 0 | 0 | — |
case-11 | pass→pass | 7,606 | 6,614 | -13% | 1 | 1 | 0% | 1,292 | 2,655 | +105% | 0 | 0 | — |
case-12 | pass→pass | 12,044 | 12,532 | +4% | 1 | 1 | 0% | 1,893 | 3,727 | +97% | 0 | 0 | — |
case-13 | pass→pass | 12,874 | 9,304 | -28% | 1 | 1 | 0% | 2,033 | 3,068 | +51% | 0 | 0 | — |
case-14 | pass→pass | 16,711 | 15,022 | -10% | 1 | 1 | 0% | 2,772 | 3,992 | +44% | 0 | 0 | — |
case-15 | pass→pass | 13,575 | 12,339 | -9% | 1 | 1 | 0% | 2,240 | 3,666 | +64% | 0 | 0 | — |
case-16 | pass→pass | 12,483 | 9,983 | -20% | 1 | 1 | 0% | 2,007 | 3,184 | +59% | 0 | 0 | — |
case-17 | pass→pass | 11,416 | 9,586 | -16% | 1 | 1 | 0% | 1,788 | 3,254 | +82% | 0 | 0 | — |
case-18 | pass→pass | 12,520 | 8,221 | -34% | 1 | 1 | 0% | 1,912 | 2,916 | +53% | 0 | 0 | — |
case-19 | fail→pass | 13,341 | 10,581 | -21% | 1 | 1 | 0% | 1,959 | 3,219 | +64% | 0 | 0 | — |
case-20 | pass→pass | 12,204 | 6,341 | -48% | 1 | 1 | 0% | 1,789 | 2,504 | +40% | 0 | 0 | — |
case-21 | fail→pass | 13,488 | 3,961 | -71% | 1 | 1 | 0% | 2,006 | 2,172 | +8% | 0 | 0 | — |
case-22 | pass→pass | 22,596 | 17,897 | -21% | 1 | 1 | 0% | 3,059 | 4,107 | +34% | 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 +27 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.