WILMINGTON, Del., June 9, 2026 — SpiderPublish™, the content platform from SpiderIQ™, serves any tenant's published documentation as a Model Context Protocol (MCP) server from a single endpoint, https://spideriq.ai/api/v1/content/docs/mcp, available today, so that an AI agent can read a product's current documentation instead of a copy it memorised months earlier. A caller names the site in an X-Content-Domain header; there is no per-site server, SDK or deployment step involved.

"An agent answering questions about your product is usually reading documentation it memorised months ago. That is not fixed by a better model. It is fixed by giving the agent somewhere current to look," said Martin Shein, Founder of SpiderIQ.

The endpoint speaks JSON-RPC 2.0 over Streamable HTTP, the transport MCP clients already use, and exposes four read-only tools scoped to a single site's published pages:

  • search_docs — full-text search across the site's published documentation. Live for every site.

  • get_doc — fetch one published page as Markdown by path. Live for every site.

  • semantic_search_docs and ask_docs — vector search and a cited answer. Both return an error flagged temporarily unavailable until Docs AI is enabled for the site, according to SpiderIQ's own verification record.

In a check across five site domains, per SpiderIQ's own verification record, each domain returned results drawn only from its own published pages, and a domain with no published documentation returned none. Those figures were measured by SpiderIQ rather than by an outside party; the method was the same JSON-RPC call issued once per domain with only the X-Content-Domain header changed.

The endpoint reads the published documentation record directly rather than a separate search index, so a page becomes queryable when it is published and there is no index to rebuild or reconcile.

"Two of the four tools work everywhere today and two are gated on a plan. I would rather tell you which half is live than let a developer find out from an agent," said Shein.

SpiderIQ has no external adopter to name for the feature yet. The endpoint is public for published content and self-describing: a tools/list call returns each tool's name, description and input schema, so a client can discover the surface without separate documentation. The capability is described at https://publish.spideriq.ai/docs-platform.

About SpiderPublish

SpiderPublish is an AI-native headless CMS from SpiderIQ, built multi-tenant and multi-project so an agency can use any agent runtime to build and run client sites: websites, funnels, interactive forms, blogs, API documentation with AI-RAG search, image and file hosting, a changelog and a newsroom, across one or many domains on a single deploy. It brings vibe publishing to any agency. Learn more at https://publish.spideriq.ai.

Media contact

Martin Shein Founder, SpiderIQ martin@spideriq.ai https://publish.spideriq.ai/press