Skip to content

Model Context Protocol endpoint · read-only · no key required

adaptation.ai, for agents

Our site is agent-consumable.

adaptation.ai exposes a read-only endpoint over the open Model Context Protocol (MCP). Any MCP-compatible agent or assistant can connect and query what we offer, what we have written, and how to reach us, without scraping a single page.

What this is

MCP is an open standard for connecting AI agents to tools and data. An agent that speaks MCP can discover our tools and call them the same way it connects to any other MCP server. We built the endpoint because we are an applied AI agency, and a site our own agents can read is the honest version of that.

Endpoint

POST https://www.adaptation.ai/api/mcp

Transport: Streamable HTTP. Access: read-only, no authentication. Point any MCP client at the URL and run the standard handshake (initialize → list tools → call a tool).

What you can do

Five read-only tools (descriptions are the live tool strings, verbatim):

ToolWhat it returnsInput
get_offeringsReturn Adaptation AI's service offerings: the Educate, Build, and Run lines, the commercial formats, and the Solutions products.none
search_insightsSearch Adaptation AI's published insights by keyword and return matching titles, summaries, and links.{ query: string }
get_faqReturn Adaptation AI's frequently asked questions and answers, optionally filtered by a search term.{ query?: string }
get_booking_linkReturn the link to book a scoping call with Adaptation AI.none
ask_adaptationAsk a question about Adaptation AI and get an answer grounded only in its published site content.{ question: string }

The boundaries (so you know what to expect)

  • Every tool is read-only. There is nothing to write, nothing to change, no account to hold.
  • Tools serve only our published, public content. No personal data, no internal material.
  • Requests are rate-limited and capped, so the endpoint stays fast for everyone.
  • Responses are plain facts and links. If you need a person, get_booking_link returns the way to reach one.

Why we did it

We help teams put AI to work: we teach them, we build the tool, and we run it in production. Making our own site agent-consumable is the same idea applied to us. It also means agents and answer engines can find what we do directly, which is where more of the web is heading.