Documentation Index
Fetch the complete documentation index at: https://mintlify.com/AgricIDaniel/claude-seo/llms.txt
Use this file to discover all available pages before exploring further.
Claude SEO’s link and clustering commands give you two complementary ways to understand authority and content architecture. /seo backlinks dissects the inbound link profile using a free-to-premium data cascade that starts with Common Crawl and scales up through Moz and Bing Webmaster to DataForSEO. /seo cluster uses SERP overlap — not text similarity — to group keywords into hub-and-spoke content clusters and generates an interactive visualization. /seo sxo closes the loop by analyzing whether a page’s type actually matches what Google rewards for its target keyword.
/seo backlinks
/seo cluster
/seo sxo
/seo backlinks <url>
/seo backlinks runs a seven-section backlink profile against a 3-tier data cascade. No API keys are required to start — Common Crawl domain metrics are always available. Adding Moz and Bing Webmaster (both free with signup) unlocks domain authority, anchor text distribution, and competitive comparison. The DataForSEO extension unlocks live link velocity and new/lost tracking.# Full profile using all configured sources
/seo backlinks https://example.com
# Competitor backlink gap analysis
/seo backlinks gap https://example.com https://competitor.com
# Toxic link detection and disavow recommendations
/seo backlinks toxic https://example.com
# New and lost backlinks (DataForSEO only)
/seo backlinks new https://example.com
# Verify known backlinks still resolve
/seo backlinks verify https://example.com --links links.txt
# Setup instructions for free APIs
/seo backlinks setup
Data sources
| Source | What it provides | Confidence |
|---|
| DataForSEO (premium) | Full profile, velocity, new/lost, spam score | 1.0 |
| Moz API (free signup) | Domain Authority, Page Authority, Spam Score, anchors | 0.85 |
| Bing Webmaster Tools (free signup) | Link details, anchor text, competitor comparison | 0.70 |
| Common Crawl (always available) | In-degree, PageRank, harmonic centrality | 0.50 |
| Verification Crawler (always available) | Live link resolution check | 0.95 |
What it analyzes
Profile overview: Total referring domains, follow ratio, domain diversity (flag if any single domain exceeds 5% share), and trend direction (growing / stable / declining).Anchor text distribution: Healthy benchmarks are 30–50% branded, 15–25% naked URL, 10–20% generic, 3–10% exact match, 5–15% partial match. Exact-match anchors exceeding 15% are flagged as a Google Penguin risk signal.Referring domain quality: TLD distribution (.edu, .gov, .org = high authority; excessive .xyz, .info = low quality), country distribution vs target market, domain rank distribution across authority tiers.Toxic link detection: 30 toxic patterns in the reference file, including known PBN domains, 100% exact-match anchor from a single domain, link farms (10K+ outbound links per page), and paid link patterns (footer/sidebar links across all pages of a domain). Moz Spam Score >11% (on a 1–17 scale) triggers a high-risk flag.Competitor gap analysis: Bing Webmaster is the only free tool with built-in competitor comparison. The output lists domains linking to competitor but not target (link building opportunities), domains linking to both (validate existing relationships), and domains linking only to target (competitive advantage).Health score
The Backlink Health Score (0–100) requires at least 4 of 7 factors to have a data source. If fewer than 4 factors have data, the report displays INSUFFICIENT DATA (X/7 factors scored) instead of a numeric score to avoid misleading results. When only Common Crawl is available, the maximum score is capped at 70/100./seo cluster <seed-keyword>
/seo cluster was contributed to Claude SEO by Lutfiya Miller (Pro Hub Challenge Winner, v1.9.0). It groups keywords by how Google actually ranks them — through shared top-10 SERP results — rather than by text similarity, producing a hub-and-spoke content architecture with an internal link matrix and an interactive cluster-map.html visualization.# Full planning workflow: expand → cluster → architect → visualize
/seo cluster plan "claude code skills"
# Import keyword set from an existing /seo plan output
/seo cluster plan --from strategy
# Execute plan: create content via claude-blog or write briefs
/seo cluster execute
# Regenerate the interactive cluster visualization
/seo cluster map
How SERP-overlap clustering works
Starting from a seed keyword, the skill expands to 30–50 variants using related searches, People Also Ask questions, long-tail modifiers, and intent variations. For each candidate keyword pair, it fires a SERP search and counts shared URLs in the top 10 organic results:| Shared results | Relationship | Action |
|---|
| 7–10 | Same post | Merge into single target page |
| 4–6 | Same cluster | Group under same spoke cluster |
| 2–3 | Interlink | Adjacent clusters, add cross-links |
| 0–1 | Separate | Different clusters or exclude |
Full pairwise comparison for 40 keywords = 780 checks. The skill pre-groups by intent (reducing this to ~180), only cross-checks group boundary keywords, and skips obvious long-tail variants of the same head term.What it produces
- Cluster plan (
cluster-plan.json + cluster-plan.md): Pillar keyword, cluster groupings, spoke posts per cluster, intent classification (informational / commercial / transactional), content templates (ultimate-guide, how-to, listicle, comparison, best-of, landing page), and word count targets (pillar: 2,500–4,000 words; spokes: 1,200–1,800 words).
- Internal link matrix: Every spoke links to the pillar (mandatory), pillar links to every spoke (mandatory), spoke-to-spoke within cluster (2–3 per post), cross-cluster (0–1 per post). No orphan pages.
- Interactive visualization (
cluster-map.html): An SVG cluster map with status tracking (planned / written) that updates after /seo cluster execute.
- Content briefs (
cluster-briefs/): Full H2/H3 outlines, secondary keywords, and anchor text targets — generated when claude-blog is not installed.
Cannibalization guard
No two posts may share the same primary keyword. If SERP overlap between two planned pages reaches 7+, they are automatically merged into a single post targeting both keywords./seo sxo <url>
/seo sxo (Search Experience Optimization) was contributed by Florian Schmitz (Pro Hub Challenge, v1.9.0). It reads the SERP backwards: instead of auditing technical compliance, it asks whether the page’s type matches what Google is actually rewarding for the target keyword. A page can score 95/100 on technical SEO and still fail to rank because it is the wrong page type.# Full SXO analysis — auto-detects keyword from title + H1
/seo sxo https://example.com/blog/how-to-x
# Specify the target keyword explicitly
/seo sxo https://example.com/blog/how-to-x "target keyword"
# Generate IST/SOLL wireframe with concrete section-level placeholders
/seo sxo wireframe https://example.com/blog/how-to-x
# Persona scoring only (skip SERP analysis)
/seo sxo personas https://example.com/blog/how-to-x
Page-type mismatch detection
The skill classifies the target page and each of the top-10 SERP results using a shared page-type taxonomy, then checks for consensus (>60% = strong, 40-60% = mixed, <40% = fragmented):| Your page type | SERP expects | Severity |
|---|
| Blog Post | Product Pages | CRITICAL |
| Blog Post | Comparison | HIGH |
| Product | Informational | HIGH |
| Landing Page | Tool/Calculator | HIGH |
| Service Page | Local Results | MEDIUM |
User story derivation
User stories are derived directly from SERP signals: People Also Ask reveals knowledge gaps, ad copy reveals commercial triggers, related searches map the journey, and the featured snippet format reveals the expected answer structure. Each story follows: As a [persona], I want to [goal], because [driver], but I'm blocked by [barrier].Multi-persona scoring
The skill derives 4–7 personas from SERP signals and scores the page against each across four dimensions: Relevance, Clarity (can this persona find their answer in 10 seconds?), Trust, and Action (is there a clear next step?). Recommendations are sorted by weakest persona first.SXO score vs SEO Health Score
The SXO Gap Score (0–100) is explicitly separate from the SEO Health Score. Both are reported when both are available: a page can score 95 SEO + 30 SXO — technically perfect but strategically misaligned.