Backlink Analysis Claude Skill
The Backlink Analysis Claude skill audits a site's link profile for SEO specialists and agencies. It reviews referring domains, anchor text, toxic signals, top pages and competitor gaps. It is free and open source (MIT), part of the claude-seo plugin by Agrici Daniel, and it works with free data sources before any paid tool.
/plugin marketplace add AgriciDaniel/claude-seo/plugin install claude-seo@agricidaniel-claude-seonpx skills add AgriciDaniel/claude-seo --skill seo-backlinksWhat the Backlink Analysis skill does
Backlink tools are expensive, and many marketers just need a sound first read of a profile. This skill detects which data sources you have and builds the report from those. Common Crawl and a link verification crawler are always available. Moz, Bing Webmaster and Keywords Everywhere need free sign-ups.
Its most distinctive rule is honesty about data. When too few of its seven scoring factors have data, it prints insufficient data rather than a health score, and with Common Crawl alone it prints not assessed instead of any number. A validator script checks the report before you see it.
The seven sections cover profile overview, anchor mix, referring domain quality, toxic links, top pages, competitor gaps and new or lost links. Some parts, such as link velocity, need the paid DataForSEO extension. The skill says so rather than faking them.
- Detects which backlink data sources are available and builds the report from them
- Checks anchor text mix against benchmark ranges and flags over-optimized exact-match anchors
- Reviews referring domain quality by TLD, country and authority tier
- Flags high and medium risk link patterns, with disavow candidates for review
- Finds top linked pages, pages with no links and broken URLs that still have links
- Compares your domain with competitors to list link opportunities
- Verifies whether known backlinks still exist and marks JavaScript pages as unverifiable
- Produces a 0 to 100 health score only when enough factors have real data
How it works
Detect sources
Checks for DataForSEO, Moz, Bing Webmaster and Keywords Everywhere, and always includes Common Crawl and the verification crawler.
Analyze seven areas
Covers profile overview, anchors, referring domains, toxic links, top pages, competitor gaps and new or lost links, using the best source for each.
Score with confidence
Weights each factor by source quality. If fewer than four factors have data, it reports insufficient data, not a score.
Validate the report
Runs a validator that fails any report with numbers it has no scoreable source for, then fixes it.
Review and report
Fact-checks every claim and source label, then lists critical, high and medium issues and the top link-building opportunities.
Example prompts
Once the Backlink Analysis skill is installed, ask Claude in plain language. These prompts trigger it:
- Analyze the backlink profile of example.com
- Run a backlink gap analysis between our site and competitor.com
- Find toxic backlinks on our domain and list disavow candidates
- Check whether these 40 backlinks I earned are still live
- What free API keys should I set up for backlink data?
What you get
- A backlink report with section statuses, source labels and a health score or insufficient data notice
- Critical, high and medium priority issue lists
- A top-10 link-building opportunity list
- A competitor link gap list
- A verification result for known links
Who it is for
- SEO specialists who want a first-pass link audit without paying for a big tool
- Agencies auditing new clients
- Site owners checking for spammy links
- Teams verifying that earned links are still live
How to install the Backlink Analysis skill
What you need: Claude Code with the claude-seo plugin and Python 3.11 or newer. Free Moz and Bing keys give richer data. Link velocity needs the paid DataForSEO extension.
- Claude Code plugin
/plugin marketplace add AgriciDaniel/claude-seo - Claude Code plugin
/plugin install claude-seo@agricidaniel-claude-seo - Skills CLI
npx skills add AgriciDaniel/claude-seo --skill seo-backlinks - Manual
git clone --depth 1 https://github.com/AgriciDaniel/claude-seo.git && bash claude-seo/install.sh
After installing, restart Claude Code or start a new session. Claude loads the Backlink Analysis skill automatically when your request matches it, or you can ask for it by name.
Tips for better results
- Run the setup command first and add free Moz and Bing keys. Common Crawl alone gives only rank signals.
- Treat flagged toxic links as review items. Check them manually before you disavow anything.
- Bing comparisons only work for properties you have registered, so use other sources for arbitrary competitors.
Limitations
- Free sources give point-in-time snapshots. New and lost link tracking needs the paid DataForSEO extension.
- With only Common Crawl you get no numeric scores by design, so the report is thinner.
- It relies on plugin scripts, so installing just the SKILL.md folder with the Skills CLI will not run them.
Backlink Analysis Claude skill FAQ
Yes. It is open source under the MIT license, published by Agrici Daniel on GitHub. You can read, copy and modify it at no cost. Any paid data tools it can connect to are optional and billed separately by those vendors.
Run /plugin marketplace add AgriciDaniel/claude-seo, then /plugin install claude-seo@agricidaniel-claude-seo. Run /seo backlinks setup to add free API keys, then /seo backlinks followed by your URL. The manual installer is the other route if you skip plugins.
Claude Code is the safer choice. This skill calls plugin scripts for data lookups, the link crawler and the report validator, which need a terminal. You can still zip the SKILL.md and upload it under Settings, Capabilities, Skills on a paid plan, but expect reduced results.
Partly. With this skill, Claude can use Common Crawl, a link verification crawler and free Moz or Bing keys to review a profile. Coverage is thinner than paid indexes. Link velocity and new or lost link tracking need a paid DataForSEO extension.
It checks for patterns such as private blog network signals, over-matched anchors, mass directory links, link farms and sitewide paid-style links. It treats them as review flags, not proof, and lists disavow candidates for you to confirm before acting.