<?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: GrowNexus</title>
    <description>The latest articles on DEV Community by GrowNexus (@grownexus).</description>
    <link>https://dev.to/grownexus</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%2F4132435%2F7df8d4e2-5841-4537-a1ea-07433c1bcb75.png</url>
      <title>DEV Community: GrowNexus</title>
      <link>https://dev.to/grownexus</link>
    </image>
    <atom:link rel="self" type="application/rss+xml" href="https://dev.to/feed/grownexus"/>
    <language>en</language>
    <item>
      <title>We Audited 25 Global SaaS Platforms for llms.txt v2: Here is What We Found</title>
      <dc:creator>GrowNexus</dc:creator>
      <pubDate>Sat, 19 Sep 2026 06:09:55 +0000</pubDate>
      <link>https://dev.to/grownexus/we-audited-25-global-saas-platforms-for-llmstxt-v2-here-is-what-we-found-218o</link>
      <guid>https://dev.to/grownexus/we-audited-25-global-saas-platforms-for-llmstxt-v2-here-is-what-we-found-218o</guid>
      <description>&lt;p&gt;When the &lt;code&gt;llms.txt&lt;/code&gt; proposal emerged as a standard to help LLMs and autonomous agents navigate web documentation efficiently, it promised an alternative to messy HTML scraping.&lt;/p&gt;

&lt;p&gt;In August 2026, the specification was formally updated to &lt;strong&gt;v2&lt;/strong&gt;, introducing two major changes:&lt;/p&gt;

&lt;ol&gt;
&lt;li&gt;Standardized discovery via HTTP headers and HTML &lt;code&gt;&amp;lt;link rel="describedby"&amp;gt;&lt;/code&gt; tags.&lt;/li&gt;
&lt;li&gt;Markdown twin conventions for individual documentation routes.&lt;/li&gt;
&lt;/ol&gt;

&lt;p&gt;Our research team at &lt;strong&gt;&lt;a href="https://grow-nexus.com" rel="noopener noreferrer"&gt;GrowNexus&lt;/a&gt;&lt;/strong&gt; audited 25 major SaaS platforms to measure how the industry is actually adopting the standard.&lt;/p&gt;




&lt;h2&gt;
  
  
  📊 Key Findings from the Dataset
&lt;/h2&gt;

&lt;p&gt;We examined 25 prominent platforms across developer tooling, CMS, infrastructure, and productivity software:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;strong&gt;Adoption Rate:&lt;/strong&gt; &lt;strong&gt;18 of 25 (72%)&lt;/strong&gt; publish an &lt;code&gt;llms.txt&lt;/code&gt; file.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;File Size Variance:&lt;/strong&gt; File sizes ranged from &lt;strong&gt;2,088 bytes to 210,736 bytes&lt;/strong&gt; (median: &lt;strong&gt;16,299 bytes&lt;/strong&gt;).&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Spec Structure Compliance:&lt;/strong&gt; Only &lt;strong&gt;12 of 18 (66.7%)&lt;/strong&gt; follow the section hierarchy (H1, blockquote summary, curated H2 list) outlined in the spec.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Header Discoverability (&lt;code&gt;rel="describedby"&lt;/code&gt;):&lt;/strong&gt; Only &lt;strong&gt;2 of 18 (11.1%)&lt;/strong&gt; emit the header.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;HTML-Level Discoverability:&lt;/strong&gt; &lt;strong&gt;0 of 18 (0%)&lt;/strong&gt; implemented &lt;code&gt;&amp;lt;link rel="describedby"&amp;gt;&lt;/code&gt; in their page markup.&lt;/li&gt;
&lt;/ul&gt;




&lt;h2&gt;
  
  
  ⚠️ The Problem with 100KB+ &lt;code&gt;llms.txt&lt;/code&gt; Files
&lt;/h2&gt;

&lt;p&gt;The most significant operational issue discovered is using &lt;code&gt;llms.txt&lt;/code&gt; as a raw sitemap dump. &lt;/p&gt;

&lt;p&gt;When autonomous agents (like Claude Code, Codex, or Hermes) parse context, scratchpads are token-budgeted. Files over 50KB consume 15,000+ tokens before the agent executes a single tool call, resulting in aggressive truncation.&lt;/p&gt;

&lt;p&gt;A well-architected &lt;code&gt;llms.txt&lt;/code&gt; should act as a &lt;strong&gt;curated index&lt;/strong&gt; of core endpoints—not a full site mirror.&lt;/p&gt;




&lt;h2&gt;
  
  
  🛠️ Open-Source Benchmark Dataset
&lt;/h2&gt;

&lt;p&gt;We have released the complete dataset and analysis scripts under the CC BY 4.0 license:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;strong&gt;GitHub Repository:&lt;/strong&gt; &lt;a href="https://github.com/grownexus/llms-txt-benchmark" rel="noopener noreferrer"&gt;github.com/grownexus/llms-txt-benchmark&lt;/a&gt;
&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Complete Implementation Guide &amp;amp; Code:&lt;/strong&gt; &lt;a href="https://grow-nexus.com/insights/llms-txt-guide/" rel="noopener noreferrer"&gt;GrowNexus Insights&lt;/a&gt;
&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;How is your engineering team approaching &lt;code&gt;llms.txt&lt;/code&gt; implementation for your APIs?&lt;/p&gt;

</description>
      <category>agents</category>
      <category>ai</category>
      <category>llm</category>
      <category>saas</category>
    </item>
  </channel>
</rss>
