MCP Server

Driive includes a built-in MCP (Model Context Protocol) server that lets AI agents interact with your scheduling data. Instead of building custom API integrations for every AI tool, MCP provides a standardized way for AI assistants to check availability, create appointments, look up customers, and more — all through a single protocol.

Note: The MCP Server is in beta. Available tools, names, and inputs are subject to change as we refine the integration — avoid relying on it for production-critical automations.

What is the Model Context Protocol?

The Model Context Protocol (MCP) is an open standard for connecting AI agents to external tools and data sources. Think of it as a universal adapter: any AI tool that supports MCP can plug into Driive without custom integration code.

What this means for you:

AI assistants like Claude can read and manage your Driive data
Voice-to-booking workflows can check availability and create appointments in real time
Custom AI agents can automate scheduling tasks on behalf of your team
You get a standard, secure interface instead of building one-off API integrations

Use cases

Scenario	How it works
AI scheduling assistant	A customer interacts with a chatbot that checks your real-time availability and books an appointment — no human intervention needed
Voice-to-booking	A customer calls your business, speaks with an AI voice assistant, and the assistant creates the appointment in Driive
Automated appointment management	An AI agent monitors your schedule and proactively reschedules or fills open slots based on rules you define
Internal team assistant	Your dispatcher asks an AI assistant "What's on the schedule for tomorrow?" and gets an instant answer pulled from Driive
Smart lead qualification	An AI agent reviews incoming appointment requests and triages them based on appointment type, location, and availability

Setting up the MCP Server

Step 1: Navigate to the MCP Server page

Go to Setup > Developer > MCP Server in the sidebar.

Step 2: Copy your server endpoint URL

The MCP Server page displays your organization's unique endpoint URL. This is the URL your AI agent will connect to. Copy it.

Step 3: Create an API key for authentication

MCP connections are authenticated with an API key, passed in the Authorization header. If you don't already have one:

Navigate to Setup > Developer > API Keys.
Create a new key with a descriptive name (e.g., "MCP Connection").
Copy the key — it's shown only once.

See API Keys for detailed instructions.

Step 4: Configure your AI agent

In your AI tool or agent framework, add Driive as an MCP server using:

Endpoint URL — The URL you copied in Step 2
API Key — The key you created in Step 3

The exact configuration steps depend on your AI tool. Consult your tool's documentation for how to add an MCP server connection.

Available capabilities

Once connected, your AI agent can use the following capabilities through the MCP server.

Tools

Tools are actions the AI agent can perform. The MCP Server page shows the live list — every tool's name and description — in the Tools table. They cover things like:

Find appointments — Look up existing appointments and their details
Check availability — Search open time slots for an appointment type
Create service requests — Log a new lead with what the customer needs
Book appointments — Schedule appointments with customer details, appointment type, and time
Manage appointments — Update, reschedule, or cancel existing appointments
Work with customers — Find or create customer records
Query appointment types — List available services
Send booking links — Share a link so the customer picks their own time

Because the MCP Server is in beta, check the Tools table on the page for the current, authoritative list.

Prompts

The MCP server also exposes pre-built prompts that help AI agents understand how to interact with your Driive data effectively. These prompts provide context about your organization's appointment types, scheduling rules, and data structure.

Security

MCP connections are secured with the same safeguards as the rest of Driive's developer tools:

API key authentication — Every MCP request must include a valid API key
Organization-scoped — The AI agent can only access data belonging to your organization
Audit trail — MCP requests are tracked the same way as API requests, including last-used timestamps on your API key

Security recommendations

Use a dedicated API key for your MCP connection — don't share it with other integrations, so you can revoke it independently
Revoke the key immediately if you decommission the AI agent or suspect unauthorized access

Troubleshooting

If your AI agent can't connect to the MCP server or isn't returning expected results:

Verify the endpoint URL — Make sure you copied the full URL from the MCP Server page
Check the API key — Confirm the key is still listed on the API Keys page (a revoked key stops working immediately) and is sent as Authorization: Bearer <your-api-key>
Test with a simple query — Start with a read-only operation (like listing appointment types) to confirm the connection works before trying more advanced actions
Check your AI tool's MCP documentation — Configuration varies by tool; make sure the endpoint and authentication are set up correctly

For more detailed troubleshooting, see Troubleshooting Developer Tools.

Next steps

API Keys — Create and manage the API key that authenticates your MCP connection
Developer Overview — See how MCP fits into Driive's full developer platform
Webhooks — Combine MCP with webhooks for a complete integration (e.g., AI agent responds to real-time events)