<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:dc="http://purl.org/dc/elements/1.1/">
  <channel>
    <title>DEV Community: OPENSEO</title>
    <description>The latest articles on DEV Community by OPENSEO (@openseo).</description>
    <link>https://dev.to/openseo</link>
    <image>
      <url>https://media2.dev.to/dynamic/image/width=90,height=90,fit=cover,gravity=auto,format=auto/https:%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Fuser%2Fprofile_image%2F4159494%2F81e769ad-10ce-4d89-a428-5bd267bd6fa6.png</url>
      <title>DEV Community: OPENSEO</title>
      <link>https://dev.to/openseo</link>
    </image>
    <atom:link rel="self" type="application/rss+xml" href="https://dev.to/feed/openseo"/>
    <language>en</language>
    <item>
      <title>I built a free backlink intelligence skill for AI agents (no Ahrefs subscription needed)</title>
      <dc:creator>OPENSEO</dc:creator>
      <pubDate>Sat, 03 Oct 2026 19:26:54 +0000</pubDate>
      <link>https://dev.to/openseo/i-built-a-free-backlink-intelligence-skill-for-ai-agents-no-ahrefs-subscription-needed-1nlh</link>
      <guid>https://dev.to/openseo/i-built-a-free-backlink-intelligence-skill-for-ai-agents-no-ahrefs-subscription-needed-1nlh</guid>
      <description>&lt;p&gt;&lt;strong&gt;Backlink data shouldn't cost $99/month.&lt;/strong&gt; So I built &lt;code&gt;backlink-radar&lt;/code&gt; — an open-source skill that teaches any AI agent (Claude, Cursor, Kimi, whatever you run) to answer "who links to this site?" using data sources that cost $0.&lt;/p&gt;

&lt;h2&gt;
  
  
  The problem
&lt;/h2&gt;

&lt;p&gt;Every SEO guide says "check your competitors' backlinks." Then you open Ahrefs or Semrush and hit a paywall before you've learned anything. For side projects and early-stage sites, that's a tax on curiosity.&lt;/p&gt;

&lt;p&gt;Meanwhile, a huge amount of link data is sitting in the open:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;strong&gt;Common Crawl&lt;/strong&gt; — the largest public web crawl, with a free web-graph dataset&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Google Search Console&lt;/strong&gt; — your own link data, already free&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Open PageRank&lt;/strong&gt; — free domain authority signals&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;The data is there. What's missing is the plumbing. That's what the skill is.&lt;/p&gt;

&lt;h2&gt;
  
  
  What it does
&lt;/h2&gt;

&lt;p&gt;Drop the folder into your agent's skills directory (or paste &lt;code&gt;SKILL.md&lt;/code&gt; into a system prompt), and ask:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;"Who links to stripe.com? Give me referring domains."&lt;/li&gt;
&lt;li&gt;"Here's my GSC links export — which pages earn my links?"&lt;/li&gt;
&lt;li&gt;"Link-gap analysis: my site vs two competitors."&lt;/li&gt;
&lt;li&gt;"What should I build to earn backlinks in the dev-tools niche?"&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;It also ships a &lt;strong&gt;19-check on-page SEO + GEO audit&lt;/strong&gt; (title/meta/canonical/schema, plus AI-engine readiness: robots.txt AI-bot stance, llms.txt, SSR content) — zero API keys required.&lt;/p&gt;

&lt;h2&gt;
  
  
  Radical honesty about free data
&lt;/h2&gt;

&lt;p&gt;Free sources give you &lt;strong&gt;observed&lt;/strong&gt; links, not a complete index. Common Crawl lags weeks to months. Dofollow vs nofollow can't be confirmed from crawl data alone. The skill states these limits in its own output instead of pretending to be Ahrefs.&lt;/p&gt;

&lt;p&gt;When you outgrow free, the upgrade path is pay-per-call (DataForSEO) — not a subscription you forget to cancel.&lt;/p&gt;

&lt;h2&gt;
  
  
  The part that turned into a product
&lt;/h2&gt;

&lt;p&gt;Finding who links where is step one. &lt;strong&gt;Actually getting the link&lt;/strong&gt; is the part everyone hates. So alongside the free skill, I run the &lt;strong&gt;Backlink Hub&lt;/strong&gt; — a member directory of sites that accept guest posts, resource listings, and niche edits, with an MCP server your agent can query directly:&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight json"&gt;&lt;code&gt;&lt;span class="p"&gt;{&lt;/span&gt;&lt;span class="w"&gt;
  &lt;/span&gt;&lt;span class="nl"&gt;"mcpServers"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="p"&gt;{&lt;/span&gt;&lt;span class="w"&gt;
    &lt;/span&gt;&lt;span class="nl"&gt;"backlink-hub"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="p"&gt;{&lt;/span&gt;&lt;span class="w"&gt;
      &lt;/span&gt;&lt;span class="nl"&gt;"type"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="s2"&gt;"http"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;&lt;span class="w"&gt;
      &lt;/span&gt;&lt;span class="nl"&gt;"url"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="s2"&gt;"https://hub.recmoment.net/mcp"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;&lt;span class="w"&gt;
      &lt;/span&gt;&lt;span class="nl"&gt;"headers"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="p"&gt;{&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="nl"&gt;"Authorization"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="s2"&gt;"Bearer rmh_YOUR_KEY"&lt;/span&gt;&lt;span class="w"&gt; &lt;/span&gt;&lt;span class="p"&gt;}&lt;/span&gt;&lt;span class="w"&gt;
    &lt;/span&gt;&lt;span class="p"&gt;}&lt;/span&gt;&lt;span class="w"&gt;
  &lt;/span&gt;&lt;span class="p"&gt;}&lt;/span&gt;&lt;span class="w"&gt;
&lt;/span&gt;&lt;span class="p"&gt;}&lt;/span&gt;&lt;span class="w"&gt;
&lt;/span&gt;&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;Free tier: 20 directory queries/month. You can earn matching credits by submitting quality sites to the directory — or just subscribe if your time is worth more.&lt;/p&gt;

&lt;p&gt;Guardrails are in code, not in a policy page: paid placements always generate &lt;code&gt;rel="sponsored"&lt;/code&gt;, every site is vetted, and the publish decision always belongs to the publisher.&lt;/p&gt;

&lt;h2&gt;
  
  
  Links
&lt;/h2&gt;

&lt;ul&gt;
&lt;li&gt;Repo (MIT): &lt;a href="https://github.com/mingmprod-arch/backlink-radar" rel="noopener noreferrer"&gt;https://github.com/mingmprod-arch/backlink-radar&lt;/a&gt;
&lt;/li&gt;
&lt;li&gt;Free Hub API key: &lt;a href="https://recmoment.net/hub-portal/" rel="noopener noreferrer"&gt;https://recmoment.net/hub-portal/&lt;/a&gt;
&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;Feedback, issues and PRs welcome. If you run an agent framework and want the tools listed natively, open an issue — happy to add the manifest format you need.&lt;/p&gt;

</description>
      <category>seo</category>
      <category>ai</category>
      <category>opensource</category>
      <category>webdev</category>
    </item>
  </channel>
</rss>
