Cloudflare Ships an MCP Server for Its Blog — Now Any Agent Can Query Cloudflare Docs Programmatically

What shipped? Cloudflare migrated their blog to EmDash (their own CMS) and shipped an MCP server for agentic access — alongside proving Workers Cache at production scale for the first time.

What changed? The Cloudflare Blog now runs on EmDash via Cloudflare Workers behind Workers Cache — their first major production site on it. The MCP server (cloudflare-blog) exposes four tools: search_posts, list_posts, get_post, and list_tags. EmDash itself also ships an MCP server for authors to browse, create, edit, publish, and schedule content through agentic tooling. On the infrastructure side: 99.5% of static files served from cache, 70% of all requests from cache, sustained 450 RPS during Agents Week, and they absorbed a 28,000 RPS DDoS attack without blinking.

Why does a builder care? Workers Cache just got its first real-world production stamp. The EmDash MCP means agents can now manage CMS content directly — no dashboard required. And the Cloudflare Blog MCP turns 10+ years of Cloudflare's technical writing into a searchable, agent-accessible corpus. If you're building agentic workflows that need to reference Cloudflare docs, you can wire this MCP server in today and skip the scraping.