MCP server

Ask your AI assistant where things are

Connect Claude, ChatGPT, Cursor or any other MCP client to your InventAI homes. Search your inventory, browse containers, and add or move items in plain language.

https://inventaiapp.com/mcp

Free for all accounts
Sign in with InventAI, no API key
Read and write access

What you can ask

The assistant decides which InventAI tools to call. Answers include the storage path, so you know exactly where to look.

“Where did I put the camping stove?”

Tools usedlist_homes → search_items

“Add a cordless drill, $120, to the toolbox in the garage.”

Tools usedsearch_items → list_categories → create_item

“Move all winter jackets from the hallway closet to the attic storage bin.”

Tools usedbrowse_container → search_items → move_item

Connect your client

Add the server URL as a remote MCP server. The first time you use it, a browser window opens so you can sign in to InventAI and approve access.

Fastest way

Ask your agent to set it up

Claude Code, Cursor, VS Code and Windsurf agents can edit their own MCP config. Paste this prompt, then sign in when the browser opens.

Prompt

Connect the InventAI MCP server: https://inventaiapp.com/ai-assistants/

Using Claude or ChatGPT in the browser? These apps can’t change their own settings. Follow the steps below.

Or set it up manually:

Claude

Claude web, desktop and mobile apps, via custom connectors.

1Open Settings → Connectors and click “Add custom connector”.
2Enter the name InventAI and the URL https://inventaiapp.com/mcp. Leave the OAuth fields empty.
3Click Add, then Connect. Sign in to InventAI and approve access.
4In a chat, enable InventAI from the tools menu and ask about your inventory.

Team and Enterprise workspaces: an owner adds the connector first, then each member connects their own InventAI account.

Claude Code

Add the server from your terminal.

1Run the command below.
2Start Claude Code and type /mcp.
3Select inventai → Authenticate and finish sign-in in the browser.

Terminal

claude mcp add --transport http inventai https://inventaiapp.com/mcp

Add --scope user to make the server available in every project.

Cursor

Add the server to your global or project config.

1Open ~/.cursor/mcp.json (global) or .cursor/mcp.json in your project.
2Add the entry below and save.
3In Settings → MCP, click Connect next to inventai and sign in.

~/.cursor/mcp.json

{
  "mcpServers": {
    "inventai": {
      "url": "https://inventaiapp.com/mcp"
    }
  }
}

VS Code

GitHub Copilot agent mode in VS Code.

1Create .vscode/mcp.json in your workspace, or run “MCP: Add Server” from the Command Palette.
2Add the entry below and save.
3Click Start above the server entry, allow authentication and sign in.
4Open Copilot Chat in Agent mode; the InventAI tools appear in the tools list.

.vscode/mcp.json

{
  "servers": {
    "inventai": {
      "type": "http",
      "url": "https://inventaiapp.com/mcp"
    }
  }
}

Windsurf

Add the server to Cascade.

1Open Windsurf Settings → Cascade → MCP servers and click “View raw config”.
2Add the entry below and save.
3Refresh the server list and sign in when the browser opens.

~/.codeium/windsurf/mcp_config.json

{
  "mcpServers": {
    "inventai": {
      "serverUrl": "https://inventaiapp.com/mcp"
    }
  }
}

Other clients

Any client that supports remote Streamable HTTP servers with OAuth can use the URL directly.

1Add https://inventaiapp.com/mcp as a remote HTTP server. No API key or headers are needed.
2If your client only supports local (stdio) servers, use the mcp-remote bridge as shown below. It requires Node.js.
3Restart the client and sign in when the browser opens.

Generic JSON config

{
  "mcpServers": {
    "inventai": {
      "command": "npx",
      "args": [
        "-y",
        "mcp-remote",
        "https://inventaiapp.com/mcp"
      ]
    }
  }
}