Anchor Text Checker

Fetch any page and analyze its on-page anchor text. It classifies every <a href> into one of 7 anchor types, scores the mix against safe ranges, flags over-optimization patterns Google penalizes, and breaks links down by scope (internal/subdomain/external) and follow status.

This analyzes the anchors on a page you point it at, not a site's inbound backlink profile. It's a fast way to sanity-check the internal-linking and outbound-anchor pattern of a page you control or are auditing.

Built and maintained by Mojo Links, a link-building and semantic SEO agency. Hosted, no-install version: mojolinks.com/tools/anchor-text-checker.

The 7 anchor types

Each type has an independent safe range (they're per-type guidance, so they do not sum to 100%):

Type	Safe range	Example
Branded	30-50%	Mojo Links
Exact Match	2-5%	link building service
Partial Match	10-15%	best link building for agencies
Generic	15-25%	click here, read more
Naked URL	10-15%	mojolinks.com
Brand + Keyword	5-10%	Mojo Links link building
LSI / Semantic	5-15%	backlink acquisition

Install

Requires Node.js 18+.

# run without installing
npx anchor-text-checker https://example.com/post

# or install globally for the `anchor-check` command
npm install -g anchor-text-checker

Usage

# defaults: brand inferred from the domain, no target keywords
anchor-check https://example.com/post

# supply a brand and target keywords for accurate exact/partial buckets
anchor-check example.com --brand "Acme" --keywords "widgets, blue widgets"

# machine-readable output
anchor-check https://example.com --json > report.json

Example

$ anchor-check https://example.com --brand "Acme" --keywords "widgets"
Verdict: OVER-OPTIMIZED
  This profile shows over-optimization signals. Exact-match anchors are above
  the safe range, the pattern Google penalizes most.

TYPE             COUNT   YOURS   SAFE      IDEAL   STATUS
----------------------------------------------------------
Branded             8     22%    30-50%      29%    v
Exact Match         9     25%      2-5%       3%    ^
...

Issues (2):
  [CRITICAL] Exact-match anchors above the safe range
     Exact-match anchors are 25% of the profile, above the 2-5% safe range...

Options

Flag	Description
--brand <name>	Brand name for the branded / brand+keyword buckets. Defaults to the domain (example.com -> Example).
--keywords <list>	Comma-separated target keywords for the exact / partial buckets.
--json	Print the full result object as JSON.
-h, --help	Show help.

Use as a library

The scoring engine is pure and network-free; the fetch/extract layer is separate:

import { analyzeAnchors, deriveBrand } from 'anchor-text-checker'
import { toSafeUrl, fetchHtml, extractLinks } from 'anchor-text-checker/src/fetch.js'

const url = await toSafeUrl('example.com')
const { html, finalUrl } = await fetchHtml(url)
const links = extractLinks(html, finalUrl)
const result = analyzeAnchors(links, { url: finalUrl.toString(), ...deriveBrand(finalUrl.hostname), keywords: [] })

Safety

The fetcher follows redirects manually and re-checks every hop against an SSRF guard that rejects hostnames resolving to private, loopback, link-local, CGNAT, or cloud-metadata ranges. It caps the body at 2 MB and times out at 8 seconds. Only http/https pages are fetched.

Notes & limits

• Anchor classification is heuristic. Without --keywords, exact/partial buckets can't be detected, the tool will still classify branded, naked, generic, and LSI.
• Scope uses naive last-two-labels domain matching (no full public-suffix list), which is fine for typical domains.
• It reads server-rendered HTML; links injected by client-side JavaScript after load won't be seen.

License

MIT.