No Markdown representation for agents

What it detects

The site does not serve a Markdown representation of its pages via content negotiation (Accept: text/markdown) or a static .md mirror.

Why it matters

Markdown is far cheaper for an agent to parse than full HTML with its surrounding chrome and scripts; offering it directly reduces token cost and parsing errors for every agent visit.

How to fix

- Check whether the homepage responds to `Accept: text/markdown` with a Markdown representation, and whether a Markdown mirror (e.g. `/index.md`) exists as a fallback. - Implement content negotiation on `Accept: text/markdown` for at least the homepage and key content pages, returning a concise Markdown version of the page (title, summary, key links) — or publish static `.md` mirrors alongside the HTML pages. - This is the same behavior Cloudflare's "Markdown for Agents" feature provides automatically if the site sits behind Cloudflare — check whether that feature can simply be enabled instead of building it by hand. - Verify: `curl -H "Accept: text/markdown" <homepage>` and confirm a Markdown (not HTML) body is returned, or `curl` the `.md` mirror directly.

References

• RFC 9309 — Robots Exclusion Protocol
• Sitemaps protocol