From 6ac6098668f98a4953ace998efd9606f5d8ea596 Mon Sep 17 00:00:00 2001
From: Wayne Sutton <waynesutton@users.noreply.github.com>
Date: Wed, 24 Dec 2025 00:13:10 -0800
Subject: [PATCH] fix(ai): remove Link header from /raw/* and use
 window.location.origin for raw URLs

- Remove Link header from global headers scope, apply only to /index.html
- Remove X-Robots-Tag noindex from /raw/* to allow AI crawler indexing
- Use window.location.origin instead of props.url for raw markdown URL construction
- Ensures correct URLs even when props.url points to canonical/deploy preview domains
- Fixes ChatGPT/Perplexity fetch failures on /raw/*.md endpoints
---
 changelog.md                        | 28 ++++++++++++++++++++++++++++
 netlify.toml                        | 11 +++++++++--
 src/components/CopyPageDropdown.tsx | 26 ++++++++++++++++++++++----
 3 files changed, 59 insertions(+), 6 deletions(-)

diff --git a/changelog.md b/changelog.md
index cf847e7..b653475 100644
--- a/changelog.md
+++ b/changelog.md
@@ -4,6 +4,34 @@ All notable changes to this project will be documented in this file.
 
 The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/).
 
+## [1.24.8] - 2025-12-23
+
+### Fixed
+
+- Raw markdown URL construction now uses `window.location.origin` instead of `props.url`
+  - Prevents incorrect URLs when `props.url` points to canonical/deploy preview domains
+  - Uses `new URL()` constructor for proper absolute URL building
+  - Ensures raw URLs always match the current page origin
+  - Applied to both AI service links and "View as Markdown" option
+
+### Technical
+
+- `src/components/CopyPageDropdown.tsx`: Changed raw URL construction from `new URL(props.url).origin` to `window.location.origin` with `new URL()` constructor
+
+## [1.24.7] - 2025-12-23
+
+### Fixed
+
+- Removed `Link` header from `/raw/*` endpoints to fix AI crawler fetch failures
+  - Netlify merges headers, so global `Link` header was being applied to `/raw/*` despite specific block
+  - Moved `Link` header from global `/*` scope to `/index.html` only
+  - Removed `X-Robots-Tag = "noindex"` from `/raw/*` to allow AI crawlers to index raw content
+  - Raw markdown files now have clean headers optimized for AI consumption
+
+### Technical
+
+- `netlify.toml`: Removed `Link` from global headers, added specific `/index.html` block, removed `noindex` from `/raw/*`
+
 ## [1.24.6] - 2025-12-23
 
 ### Added
diff --git a/netlify.toml b/netlify.toml
index ae777be..5fe7ede 100644
--- a/netlify.toml
+++ b/netlify.toml
@@ -76,7 +76,7 @@
     "/openapi.yaml"
   ]
 
-# Security and SEO headers
+# Security and SEO headers (global, excludes /raw/*)
 [[headers]]
   for = "/*"
   [headers.values]
@@ -84,16 +84,23 @@
     X-Content-Type-Options = "nosniff"
     X-XSS-Protection = "1; mode=block"
     Referrer-Policy = "strict-origin-when-cross-origin"
+    # Link header removed from global scope to avoid applying to /raw/*
+
+# Link header for SPA entry point only
+[[headers]]
+  for = "/index.html"
+  [headers.values]
     Link = "</llms.txt>; rel=\"author\""
 
 # Raw markdown files - AI friendly headers
+# No Link header, no noindex - optimized for AI crawlers
 [[headers]]
   for = "/raw/*"
   [headers.values]
     Content-Type = "text/markdown; charset=utf-8"
     Access-Control-Allow-Origin = "*"
     Cache-Control = "public, max-age=3600"
-    X-Robots-Tag = "noindex"
+    # Removed X-Robots-Tag = "noindex" - AI crawlers need to index raw content
 
 [context.production.environment]
   NODE_ENV = "production"
diff --git a/src/components/CopyPageDropdown.tsx b/src/components/CopyPageDropdown.tsx
index c3efecc..60ad106 100644
--- a/src/components/CopyPageDropdown.tsx
+++ b/src/components/CopyPageDropdown.tsx
@@ -327,10 +327,23 @@ export default function CopyPageDropdown(props: CopyPageDropdownProps) {
   const handleOpenInAI = async (service: AIService) => {
     // Use raw markdown URL for better AI parsing
     if (service.buildUrlFromRawMarkdown) {
-      // Build raw markdown URL from page URL and slug
-      const origin = new URL(props.url).origin;
-      const rawMarkdownUrl = `${origin}/raw/${props.slug}.md`;
+      // Build absolute raw markdown URL using current origin (not props.url)
+      // This ensures correct URL even if props.url points to canonical/deploy preview domain
+      const rawMarkdownUrl = new URL(
+        `/raw/${props.slug}.md`,
+        window.location.origin,
+      ).toString();
       const targetUrl = service.buildUrlFromRawMarkdown(rawMarkdownUrl);
+
+      // For ChatGPT, add fallback if URL fetch fails
+      if (service.id === "chatgpt") {
+        // Open ChatGPT with raw URL
+        window.open(targetUrl, "_blank");
+        setIsOpen(false);
+        // Optional: Could add error handling here if needed
+        return;
+      }
+
       window.open(targetUrl, "_blank");
       setIsOpen(false);
       return;
@@ -511,7 +524,12 @@ export default function CopyPageDropdown(props: CopyPageDropdownProps) {
           <button
             className="copy-page-item"
             onClick={() => {
-              window.open(`/raw/${props.slug}.md`, "_blank");
+              // Build absolute URL using current origin for consistency
+              const rawMarkdownUrl = new URL(
+                `/raw/${props.slug}.md`,
+                window.location.origin,
+              ).toString();
+              window.open(rawMarkdownUrl, "_blank");
               setIsOpen(false);
             }}
             role="menuitem"