index4.ai Docs by

Appearance

Use index4.ai in your AI assistant ​

Connect index4.ai to your AI assistant once. After that, the assistant can search your documents whenever you ask it to — and tell you which document and page an answer came from.

You don't need to write code or copy an API key. You log in with your index4.ai account in the browser.

Pick your assistant ​

ClaudeMicrosoft Copilot via Copilot StudioGitHub Copilot (VS Code)Langdock

What you need ​

  • An index4.ai account with at least one uploaded document. No account yet? Sign up for free and upload a file in the console.
  • An AI assistant that can connect to a remote MCP server by URL, for example Claude, Microsoft Copilot Studio, GitHub Copilot in VS Code or Langdock.

Connect Claude ​

  1. In Claude, open Settings and go to Connectors.

  2. Choose Add custom connector.

  3. Enter a name, for example index4.ai, and this address:

    text
    https://mcp.index4.ai

  4. Click Connect. A browser window opens — log in with your index4.ai account and allow access.

  5. Done. index4.ai now shows up in Claude's list of connectors.

Menu names can differ

Claude's menus change from time to time and depend on your plan. If you can't find Connectors, search Claude's help for "custom connector".

Connect Microsoft Copilot (via Copilot Studio) ​

In Microsoft Copilot, index4.ai is added as a tool of an agent. Whoever builds agents in Copilot Studio does this once; colleagues then use the agent in Microsoft Copilot.

  1. In Copilot Studio, open the agent and go to Tools → Add a tool → New tool → Model Context Protocol.

  2. Enter this server address:

    text
    https://mcp.index4.ai

  3. Name it index4ai-mcp, add a short description, and choose OAuth 2.0 with dynamic discovery as the authentication.

  4. Create the connection, log in with your index4.ai account and save.

  5. Ask the agent a question that needs the tools, to test it.

Connect GitHub Copilot in VS Code ​

  1. In VS Code, open the Command Palette and run MCP: Add Server → HTTP.

  2. Enter this address:

    text
    https://mcp.index4.ai

  3. Use index4ai-mcp as the server ID and pick a scope: Global for all your projects, or Workspace for one.

  4. Sign in when the browser opens.

  5. In Copilot Chat, switch to Agent mode to use index4.ai's tools.

Connect Langdock ​

  1. In Langdock, go to Integrations → Add Integration → Start from scratch → Connect remote MCP.

  2. Enter this address:

    text
    https://mcp.index4.ai

  3. Choose OAuth (Dynamic Client Registration), click + Add connection and log in with your index4.ai account.

  4. Click Test connection, select the index4.ai tools and save (⌘/Ctrl + S).

Menu names can differ

These steps follow the index4.ai console's quick start. Microsoft and Langdock change their menus from time to time; if a name no longer matches, look for "MCP" or "remote MCP server" in the product's settings.

Try it ​

Ask your assistant in plain words. Mention index4.ai so it knows where to look:

  • "Search my index4.ai documents: what does our rental contract say about the notice period?"
  • "Using index4.ai, summarise what my documents say about the onboarding process, and tell me which file each point comes from."
  • "Add this page to index4.ai: https://example.com/handbook.pdf"

The answer includes the passages it is based on. If something looks off, open the source passage and check.

Good to know ​

  • Large files are easiest to upload in the console; your assistant can then search them right away.
  • Several projects? If you have more than one, tell the assistant which project to use.
  • Disconnect at any time in the assistant's connector settings.

Other assistants and tools ​

Any assistant that supports MCP connectors can use the same address, https://mcp.index4.ai. Setup steps for Claude Code, Cursor and other developer tools are on the MCP page.