Automation and AI

Model Context Protocol (MCP)

The Model Context Protocol (MCP) is an open standard that lets AI assistants connect to outside tools and data through a server, using one common interface.

Also called: MCP, MCP server, MCP client

An AI assistant on its own can only talk. To do something for you, such as read your submissions or create a form, it needs a way to call software. Before MCP, every assistant and every tool needed its own custom connection. MCP replaces that with one standard.

The pieces

  • MCP client. The AI app you use: Claude, ChatGPT, Cursor, VS Code and others.
  • MCP server. A program that exposes a service as a list of tools the client may call, such as “list forms” or “show spam submissions”.
  • Transport and authorisation. How the two talk, and how the user approves access. Remote servers typically use HTTP and OAuth, so you sign in and grant access in the browser.

Why it matters for tools like forms

Instead of clicking through a dashboard, you ask in plain language: “summarise this week’s enquiries”, “which submissions were marked as spam and why”, “create a contact form with name, email and message”. The assistant picks the right tool calls and reports back.

Treat the content as data

Anything an assistant reads through a tool, such as the text of a form submission, was written by a stranger. It can contain instructions aimed at the AI. A well-built setup treats such content as data to report on, never as commands to follow, and limits what the connection is allowed to change.

In StaticForm

StaticForm runs a hosted MCP server at https://api.staticform.app/mcp. It uses the Streamable HTTP transport with OAuth 2.1, and you approve read-only or read-and-write access, optionally limited to specific forms. Headless clients can use an API key instead. The tools cover forms, submissions, spam triage, delivery logs, collaboration and settings. See the MCP setup guide for steps per client.

Learn more