Azey

AI agents from ChatGPT, Claude, Perplexity, and others read your content to generate answers — but they parse through HTML, JavaScript, and navigation just to reach it. Serve Markdown gives them a direct path: clean Markdown with structured metadata. When a crawler requests it — via an <code>Accept: text/markdown</code> header or a <code>.md</code> URL — your site responds instantly. No noise, no guesswork.
<h4>How It Works</h4>
Serve Markdown adds three capabilities to your WordPress site:
1. Content Negotiation 
When an AI crawler sends <code>Accept: text/markdown</code>, your site returns Markdown instead of HTML — no URL changes needed.
2. .md URL Suffix 
Append <code>.md</code> to any post or page URL (e.g. <code>example.com/my-article.md</code>) to get the Markdown version directly.
3. Markdown Auto-Discovery 
Every page includes a <code><link rel="alternate" type="text/markdown"></code> tag in the HTML head so crawlers can find your Markdown automatically.
<h4>What the Markdown Output Looks Like</h4>
Every Markdown response includes YAML frontmatter with structured metadata followed by your post content. Here is a real example:
<pre>
---
url: 'https://akumarjain.com/textexpander-year-in-review-2025/'
title: TextExpander Year in Review 2025
author:
 name: Ajay
 url: 'https://akumarjain.com/author/akumarjain/'
date: '2026-01-05T09:05:00+05:30'
modified: '2026-01-10T14:30:06+05:30'
type: post
categories:
 - 2025
 - Year in Review
tags:
 - '#ToolsIUse'
 - Productivity
 - Remote Work
image: 'https://akumarjain.com/wp-content/uploads/text-expander-year-in-review-2025.webp'
published: true
---

# TextExpander Year in Review 2025

In my daily work, I need to type the same words and phrases repeatedly...
</pre>
<h4>Features</h4>
Content Serving
<ul>
<li>Content negotiation via <code>Accept: text/markdown</code> header</li>
<li><code>.md</code> URL suffix on any post or page</li>
<li>Auto-discovery <code><link rel="alternate"></code> tag in page head</li>
<li>Each feature can be toggled independently</li>
<li>Choose which post types are exposed (posts, pages, custom post types)</li>
</ul>
Frontmatter and Metadata
<ul>
<li>YAML frontmatter with title, author, date, categories, tags, featured image, and more</li>
<li>Toggle individual metadata fields on or off</li>
<li>Add custom static key-value pairs (e.g., <code>license</code>, <code>language</code>)</li>
<li>Map WordPress custom fields (post meta) into the frontmatter</li>
</ul>
Access Control
<ul>
<li>Per-post opt-out via a sidebar checkbox in the editor</li>
<li>Exclude entire categories from Markdown serving</li>
<li>Exclude entire tags from Markdown serving</li>
<li>Only published posts are served — drafts and private posts are never exposed</li>
</ul>
Crawler Insights
<ul>
<li>Request log showing every Markdown request with timestamp, URL, bot name, and method</li>
<li>Automatic bot detection: ClaudeBot, GPTBot, ChatGPT, OAI-SearchBot, PerplexityBot, Googlebot, Bingbot, Cohere, Meta AI, Bytespider, Applebot, and more</li>
<li>Stats dashboard: total requests, requests today, unique bots</li>
<li>Filter log by individual bot</li>
<li>Configurable retention period with automatic pruning</li>
<li>One-click log clearing</li>
</ul>
Editor Integration
<ul>
<li>“Preview Markdown” button in the post editor sidebar</li>
<li>“View Markdown” link in the Posts and Pages list table</li>
<li>Meta box shows on all enabled post types</li>
</ul>
<h4>Performance</h4>
Markdown conversion only runs when specifically requested — regular visitors see the same HTML pages they always have. The crawler log uses a lightweight custom table with three safeguards against unbounded growth: time-based retention, a row count cap, and a table size cap.
<h4>Privacy</h4>
The crawler log stores IP addresses and user-agent strings from Markdown requests. This data stays in your WordPress database and is never sent to external services. You can disable logging entirely, configure a retention period, or clear the log at any time from the settings page.

Serve Markdown

Serve Markdown versions of your posts and pages to AI agents and crawlers. Content negotiation, .md URLs, auto-discovery, and crawler logging.

Umay Ajans

Modern AI agents (ChatGPT, Claude, Perplexity, Gemini, etc.) work much better with Markdown than HTML. Umay AI Markdown inspects the incoming <code>Accept</code> header and, only when <code>text/markdown</code> is requested, intercepts the response and serves a clean, agent-friendly Markdown representation of the page.
Browsers, search engines, and any client that does not explicitly ask for Markdown receive the unchanged HTML response. There is no settings page, no cron job, and no external service call.
Key features
<ul>
<li>Zero configuration — install, activate, done.</li>
<li>Only triggers when <code>Accept: text/markdown</code> is present. Regular visitors and search engines are never affected.</li>
<li>Hybrid content extraction: uses <code>the_content</code> for posts/pages, falls back to a DOM-based extractor for archives, taxonomies, and the homepage.</li>
<li>Powered by the industry-standard <a href="https://github.com/thephpleague/html-to-markdown" rel="nofollow ugc">league/html-to-markdown</a> library.</li>
<li>Transient-cached for 12 hours per URL (sha256-keyed). Auto-invalidated on <code>save_post</code>, term edits, theme switches, and menu updates.</li>
<li>Built-in IP rate limiter (30 requests / minute by default) to mitigate abuse.</li>
<li>Strict input sanitization, header injection protection, libxml entity hardening (XXE-safe), and full WordPress Coding Standards compliance.</li>
<li>PSR-4 autoloaded, namespaced OOP code. No globals.</li>
<li>Sends <code>Vary: Accept</code>, <code>X-Robots-Tag: noindex</code>, and <code>X-Content-Type-Options: nosniff</code> on every Markdown response.</li>
</ul>
What gets sent to AI agents
Each Markdown response includes a YAML front-matter block with the page title, site name, canonical URL, and ISO-8601 generation timestamp, followed by the page body converted to Markdown. Navigation, footer, sidebars, scripts, styles, comment forms, related posts, and other page chrome are stripped before conversion.
Filters
Two filters are available for advanced customization:
<ul>
<li><code>umay_mdn_bypass</code> — Return <code>true</code> to skip Markdown handling for the current request.</li>
<li><code>umay_mdn_cache_ttl</code> — Override the default 12-hour cache lifetime (in seconds, minimum 60).</li>
<li><code>umay_mdn_rate_limit</code> — Override the default 30-requests-per-minute rate limit.</li>
<li><code>umay_mdn_converter_options</code> — Modify the league/html-to-markdown converter options array.</li>
</ul>

Umay AI Markdown

Serves clean Markdown to AI agents (ChatGPT, Claude, Perplexity) on Accept: text/markdown. HTML stays default for browsers. Zero-config.

Serve Markdown vs Umay AI Markdown