seo-backlinks

Trigger phrases

Phrases that activate this skill when typed to Claude Code:

• backlinks
• link profile
• referring domains
• toxic links
• link gap

What it does

seo-backlinks is a Claude Code skill from AgriciDaniel’s claude-seo repo. It turns Claude into a backlink analyst that works across a tiered data stack: Common Crawl and a verification crawler are always available for free; Moz API and Bing Webmaster Tools add authority scores at no cost after free signup; DataForSEO MCP unlocks the full picture if installed.

The skill produces a 7-section backlink health report: profile overview (referring domain count, follow ratio, trend), anchor text distribution with over-optimization flags, referring domain quality by TLD and country, toxic link detection with disavow criteria, top pages by backlinks, competitor gap analysis, and new/lost link velocity. It calculates a 0–100 Backlink Health Score — or honestly flags INSUFFICIENT DATA when fewer than 4 of 7 factors have a real data source.

When to use it

Reach for it when:

• You want to know whether your link profile is healthy before a content push
• You suspect negative SEO — sudden link spikes or a new batch of exact-match anchors
• You need a competitor gap list: domains linking to rivals but not to you

When not to reach for it:

• The domain is brand new with fewer than 10 backlinks; Common Crawl won’t have meaningful data yet
• You need link velocity tracking over time without DataForSEO; free sources are point-in-time snapshots only

Install

Copy the seo-backlinks SKILL.md into your project’s .claude/skills/seo-backlinks/ directory along with the supporting scripts/ folder from the repo root.

Trigger phrases: “backlinks”, “link profile”, “referring domains”, “anchor text”, “toxic links”, “link gap”, “link building”, “disavow”, “backlink audit”.

Run /seo backlinks setup the first time to configure free Moz and Bing Webmaster API keys — this takes about five minutes and meaningfully improves data confidence.

What a session looks like

A typical session has three phases:

1. Source detection. The skill checks which APIs are available — DataForSEO MCP, Moz, Bing Webmaster, Common Crawl — and reports the confidence weight for each data source before producing a single number.
2. Seven-section analysis. Profile overview, anchor distribution, domain quality, toxic link scan, top pages, competitor gap, and link velocity. Each metric is labeled with its source and confidence tier so you know what to trust.
3. Health score and action list. Critical issues first (PBN links, penalty-risk anchor ratios), then high-priority items, then link building opportunities ranked by domain authority.

The skill enforces a data sufficiency gate: if fewer than 4 of 7 scoring factors have data, it refuses to show a numeric health score and tells you exactly what to configure to fix that.

Receipts

Works well: Common Crawl catches domain-level referring count reliably for established sites — it is a useful sanity check even before you add Moz keys. The toxic link detection flags PBN patterns (100% exact-match anchor from a single domain, link farms) that are worth a manual look.

Backfires: For domains under a year old or with very low traffic, Common Crawl graph data is sparse or absent. The skill will report INSUFFICIENT DATA honestly rather than inventing a score, which means you may get that result on a real site that simply hasn’t been crawled heavily yet.

Pattern that works: Run /seo backlinks gap <url> <competitor> before any outreach campaign — the domains-linking-to-competitor-not-to-you list is the closest thing to a pre-qualified prospect list you’ll get without a paid tool.

Source and attribution

Originally written by AgriciDaniel. The canonical SKILL.md and supporting scripts live in the seo-backlinks folder of the claude-seo repository.

License: MIT. Install, adapt, and redistribute with attribution preserved.

This page documents the skill from a practitioner’s perspective. For the formal spec and updates, defer to the source repo.