Changeset 3777261

Timestamp:

10/31/2026 06:17:46 PM (8 hours ago)

Author:

umayajans

Message:

Update screenshots and captions for 1.2.0

Location:

markdown-negotiator

Files:

• assets/screenshot-1.png (modified) (1 prop) ( previous)
• assets/screenshot-2.png (modified) (1 prop) ( previous)
• assets/screenshot-3.png (modified) (1 prop) ( previous)
• assets/screenshot-4.png (added)
• assets/screenshot-5.png (added)
• tags/1.2.0/readme.txt (modified) (1 diff)
• trunk/readme.txt (modified) (1 diff)

markdown-negotiator/tags/1.2.0/readme.txt

@@ -123 +123 @@
 == Screenshots ==
 
-1. Markdown response served when an AI agent requests `Accept: text/markdown`: HTTP headers, YAML front matter and the converted Markdown body.
-2. Built-in cache in action: the first request to a URL returns `X-Cache: MISS` (Markdown generated on the fly), the next request returns `X-Cache: HIT` (served from cache, no rendering work).
-3. Transparent content negotiation: the same WordPress URL returns `text/html` to browsers and `text/markdown` to AI agents.
+1. Markdown served to an AI agent that sends `Accept: text/markdown`: HTTP headers, YAML front matter with the page's metadata, and the post as clean Markdown with its list, table and quote.
+2. Built-in cache: the first request is rendered (`X-Cache: MISS`), the next one is served from the cache (`X-Cache: HIT`). `X-Markdown-Tokens` estimates the size for the agent.
+3. Content negotiation on one URL: browsers get `text/html` with a `Link` header pointing to the Markdown version; AI agents get `text/markdown` with a canonical link.
+4. Discovery: every HTML page advertises its `.md` twin in `<head>`, and the `.md` URL returns Markdown without any special header (`noindex`, canonical link to the HTML page).
+5. The automatically generated `/llms.txt`, listing pages and posts with links to their Markdown versions and their excerpts.
 
 == Changelog ==

markdown-negotiator/trunk/readme.txt

@@ -123 +123 @@
 == Screenshots ==
 
-1. Markdown response served when an AI agent requests `Accept: text/markdown`: HTTP headers, YAML front matter and the converted Markdown body.
-2. Built-in cache in action: the first request to a URL returns `X-Cache: MISS` (Markdown generated on the fly), the next request returns `X-Cache: HIT` (served from cache, no rendering work).
-3. Transparent content negotiation: the same WordPress URL returns `text/html` to browsers and `text/markdown` to AI agents.
+1. Markdown served to an AI agent that sends `Accept: text/markdown`: HTTP headers, YAML front matter with the page's metadata, and the post as clean Markdown with its list, table and quote.
+2. Built-in cache: the first request is rendered (`X-Cache: MISS`), the next one is served from the cache (`X-Cache: HIT`). `X-Markdown-Tokens` estimates the size for the agent.
+3. Content negotiation on one URL: browsers get `text/html` with a `Link` header pointing to the Markdown version; AI agents get `text/markdown` with a canonical link.
+4. Discovery: every HTML page advertises its `.md` twin in `<head>`, and the `.md` URL returns Markdown without any special header (`noindex`, canonical link to the HTML page).
+5. The automatically generated `/llms.txt`, listing pages and posts with links to their Markdown versions and their excerpts.
 
 == Changelog ==