<?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>PromptZone - AI Prompts, Guides and Tools for Builders: LemmiLink</title>
    <description>The latest articles on PromptZone - AI Prompts, Guides and Tools for Builders by LemmiLink (@lemmilink).</description>
    <link>https://www.promptzone.com/lemmilink</link>
    <image>
      <url>https://promptzone-community.s3.amazonaws.com/uploads/user/profile_image/40718/a0963074-0127-44a6-a891-9c7f5a08d7ec.png</url>
      <title>PromptZone - AI Prompts, Guides and Tools for Builders: LemmiLink</title>
      <link>https://www.promptzone.com/lemmilink</link>
    </image>
    <atom:link rel="self" type="application/rss+xml" href="https://www.promptzone.com/feed/lemmilink"/>
    <language>en</language>
    <item>
      <title>5 prompts to audit a backlink profile with an AI assistant</title>
      <dc:creator>LemmiLink</dc:creator>
      <pubDate>Fri, 13 Nov 2026 23:01:31 +0000</pubDate>
      <link>https://www.promptzone.com/lemmilink/5-prompts-to-audit-a-backlink-profile-with-an-ai-assistant-3oco</link>
      <guid>https://www.promptzone.com/lemmilink/5-prompts-to-audit-a-backlink-profile-with-an-ai-assistant-3oco</guid>
      <description>&lt;p&gt;Most backlink audits stall on the same three questions: which links actually matter, does the anchor text look natural, and what do competitors have that I don't. These are the prompts I use with a plain CSV export from any backlink tool (the Search Console "Links" export works too). Paste the data after the prompt.&lt;/p&gt;

&lt;h2 id="1-triage-the-export"&gt;
  
  
  1. Triage the export
&lt;/h2&gt;



&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight plaintext"&gt;&lt;code&gt;You are an SEO analyst. Below is a CSV of backlinks (source URL, target URL, anchor, rel).
Group the referring domains into: editorial, directory/profile, forum/UGC, sitewide/footer, suspicious.
For each group give the count and 3 examples. Flag any domain sending more than 50 links to the same target.
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;h2 id="2-anchor-text-distribution"&gt;
  
  
  2. Anchor text distribution
&lt;/h2&gt;



&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight plaintext"&gt;&lt;code&gt;From the same data, compute the share of branded, exact-match, partial-match, naked URL and generic anchors.
Compare it with a natural profile for a site of this type and list the 5 anchors that look over-optimised.
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;h2 id="3-competitor-gap"&gt;
  
  
  3. Competitor gap
&lt;/h2&gt;



&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight plaintext"&gt;&lt;code&gt;Here are two lists of referring domains: mine and competitor X.
List the domains that link to the competitor but not to me, grouped by type (media, blog, directory, forum).
For each one, say whether a link is realistically obtainable (guest post, listing, PR) or not, and why.
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;h2 id="4-pages-to-strengthen-first"&gt;
  
  
  4. Pages to strengthen first
&lt;/h2&gt;



&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight plaintext"&gt;&lt;code&gt;Here is my Search Console export (page, query, position, impressions).
List the pages in positions 5-20 with the most impressions and rank them by how much a few new internal or external links could help.
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;h2 id="5-sanity-check-before-paying-for-a-link"&gt;
  
  
  5. Sanity check before paying for a link
&lt;/h2&gt;



&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight plaintext"&gt;&lt;code&gt;Here is a site I am considering for a sponsored article: URL, estimated traffic, main topics, and the outbound links of one sample article.
List the red flags (link-farm patterns, unrelated topics, too many outbound links per article, thin content) and give a go / no-go with reasons.
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;h2 id="running-them-on-live-data-with-mcp"&gt;
  
  
  Running them on live data with MCP
&lt;/h2&gt;

&lt;p&gt;If your assistant supports MCP, you can skip the CSV step. &lt;a href="https://lemmilink.fr" rel="nofollow ugc noopener noreferrer"&gt;LemmiLink&lt;/a&gt;, a French link-building marketplace, publishes an official MCP server with tools such as &lt;code&gt;check_existing_backlinks&lt;/code&gt;, &lt;code&gt;get_anchor_distribution&lt;/code&gt;, &lt;code&gt;find_competitors&lt;/code&gt; and &lt;code&gt;compare_sites&lt;/code&gt;, so prompts 1 to 3 can run directly from Claude.&lt;/p&gt;

&lt;p&gt;&lt;em&gt;Disclosure: this account belongs to LemmiLink.&lt;/em&gt;&lt;/p&gt;

</description>
      <category>seo</category>
      <category>prompts</category>
    </item>
  </channel>
</rss>
