Set up the MCP server

The Tuner MCP server lets your AI assistant read and change your Tuner workspace. You can configure an agent’s outcomes, intents, evals, red flags and alerts by describing what you want, and ask questions about your calls.

The Tuner MCP server is in Beta. Available actions and supported clients may change.

Why use it

Describe several outcomes, evals and intents in one message, and the assistant creates them together. Your assistant can also see your codebase, your agent’s system prompt and your call flows. Ask “based on my agent’s system prompt, suggest the right call outcomes and evals” to get a configuration that matches how your agent works.

Supported clients

Client	Auth
Claude (claude.ai)	OAuth
Claude Code	Tuner API key
Cursor	Tuner API key
Antigravity	Tuner API key

Tuner MCP server URL: https://api.usetuner.ai/mcp/

Open the setup instructions in Tuner

Tuner shows the same instructions with your IDs filled in:

While creating an agent, pick Configure via MCP in the setup wizard. See the Quickstart.
For an existing agent, click Configure with MCP at the top of Agent Settings. The dialog has Step 1: Connect to your AI client and Step 2: Start the setup, with a prompt you can copy.

The Configure with MCP dialog: IDE and chat app tabs, the Cursor, Claude Code and Antigravity options, setup steps and the server configuration to paste

IDEs

IDEs authenticate with a Tuner API key sent as a bearer token. Create one on the API Keys page; see API keys and IDs. A key is shown only when you create it.

Cursor

Open MCP settings

Go to Settings > Tools & MCP and click Add Custom MCP to open your mcp.json.

Paste the configuration

Add the following to your mcp.json:

{
  "mcpServers": {
    "tuner": {
      "type": "streamableHttp",
      "url": "https://api.usetuner.ai/mcp/",
      "headers": {
        "Authorization": "Bearer YOUR_TUNER_API_KEY"
      }
    }
  }
}

Replace YOUR_TUNER_API_KEY with your Tuner API key.

Claude Code

Run the add command

Open your terminal inside Claude Code and run:

claude mcp add tuner https://api.usetuner.ai/mcp/ --transport http --header "Authorization: Bearer YOUR_TUNER_API_KEY"

Replace YOUR_TUNER_API_KEY with your Tuner API key.

Antigravity

Open the MCP configuration

Click the … dropdown at the top of the editor’s agent panel, then Manage MCP Servers > View raw config to open mcp_config.json.

Paste the configuration

Add the following to your mcp_config.json:

{
  "mcpServers": {
    "tuner": {
      "serverUrl": "https://api.usetuner.ai/mcp/",
      "headers": {
        "Authorization": "Bearer YOUR_TUNER_API_KEY",
        "Content-Type": "application/json"
      }
    }
  }
}

Replace YOUR_TUNER_API_KEY with your Tuner API key.

Chatbots

Chatbots authenticate with OAuth. The first time you add the server, you’re redirected to Tuner to sign in and authorize access.

Claude

Add a custom connector

Go to Customize > Connectors, click the plus icon, click Add Custom Connector, and paste the Tuner MCP server URL:

https://api.usetuner.ai/mcp/

Connect and authenticate

Click Connect. You will be redirected to Tuner to sign in and authorize access.

Enable Tuner in your chat

When starting a new chat, click the + button and verify Tuner is toggled on under Connectors.

Start the setup

Once connected, paste this into your chat. Configure with MCP in Tuner gives you this prompt with your values filled in.

Set up Tuner observability for this agent using the Tuner MCP.
Fetch and follow the 'tuner_setup_agent' prompt from the Tuner MCP server.
agent_id: <AGENT_ID>
workspace_id: <WORKSPACE_ID>
agent_name: <AGENT_NAME>
direction: INBOUND

To find your workspace ID, open Workspace Settings > General Settings. Your agent’s details are on Agent Settings > Agent Setup. See API keys and IDs. You can also ask the assistant to list the agents in your workspace.

The tuner_setup_agent prompt walks your assistant through defining your call outcomes, intents, data-extraction fields, evals, red flags and alerts in one conversation.

What the MCP server can do

The server gives your assistant these tools, for the workspaces your account or key can access.

Area	What the assistant can do
Workspaces and agents	List and read workspaces and agents, and update an agent
Calls	List calls, filter them, read a call’s details, and list recently red-flagged calls
Dashboards	Read the Overview numbers and the intent and outcome breakdowns
Agent configuration	Create, update and delete call outcomes, intents, data-extraction fields, evals, capabilities, tools, red-flag and label rules, and alerts
Libraries	Read Tuner’s predefined evals and call outcome library, the agent’s analysis and voice metric settings, and the conditions available for rules and alerts

It also has prompts that guide the assistant through a task. Ask the assistant to fetch one by name:

Prompt	What it does
`tuner_setup_agent`	Configure the agent’s instructions and settings
`tuner_setup_outcomes`	Configure call success and failure outcomes
`tuner_setup_intents`	Configure call intent categories
`tuner_setup_data_fields`	Configure data-extraction fields
`tuner_setup_evals`	Configure evals
`tuner_setup_red_flags`	Configure labels and red flags
`tuner_setup_alerts`	Configure alerts
`tuner_analyze_agent`	Diagnose what is broken in the agent and recommend fixes. See Diagnose your agent.

Simulations, datasets, API keys and workspace settings are managed in the dashboard.

Example prompts

Bulk setup from a description:

“I have a restaurant booking agent. Create the following call outcomes: Reservation Booked (Success), Reservation Modified (Success), Customer Cancelled (Failure), No Availability (Failure), and General Inquiry (Success). Also add user intents for: Make a Reservation, Modify a Reservation, Cancel a Reservation, and Ask About Menu.”

Several evals at once:

“Add these evals: Did the agent confirm the reservation date and time? Did the agent ask for the party size? Did the agent offer alternative times when the requested slot was unavailable?”

Reading your current setup:

“List all agents in my workspace and show me a summary of how each one is configured.”

Best practices

Be specific, and name the agent

Detail gets better results. Instead of “add some call outcomes”, try “for my Customer Support agent, add a call outcome called ‘Appointment Booked’ with type Success, defined as: the customer scheduled an appointment with a confirmed date and time.” If you have multiple agents, always say which one.

Use bulk operations for initial setup

Describe all your outcomes, intents, and evals in one message. The assistant creates them together.

Review changes after making them

Ask your assistant to list the current settings afterwards, so you can verify everything looks right before calls start flowing.

Use the dashboard for monitoring

Use the MCP server for configuration and diagnosis. Track live results and trends over time in the Tuner dashboard.

Troubleshooting

Connection failed or timed out

Chatbots: Check that you’re using the correct MCP server URL and that your Tuner account is active.
IDEs: Check that your API key is valid and correctly placed in the configuration file. Make sure https://api.usetuner.ai/mcp/ is reachable.

Authentication error

Chatbots: Disconnect and reconnect. In Claude, go to Customize > Connectors, remove Tuner, then add it again.
IDEs: Check that the key hasn’t been revoked on the API Keys page. If you no longer have it, create a new key and update your configuration file.

Tuner not appearing in chat

Claude: Click the + button and check that Tuner is toggled on under Connectors.
IDEs: Verify the MCP server configuration is saved correctly and restart the IDE if needed.

Next steps

Diagnose your agent

Once calls are flowing, use the MCP server to find what’s broken, understand why, and fix it.

Call outcomes and intents

What you’re configuring when you set these up through MCP.

Getting Started

Integrations

Evaluation

Observability

Simulation

Datasets

MCP

Guides

Resources

Why use it

Supported clients

Open the setup instructions in Tuner

IDEs

Cursor

Claude Code

Antigravity

Chatbots

Claude

Start the setup

What the MCP server can do

Example prompts

Best practices

Troubleshooting

Next steps

Diagnose your agent

Call outcomes and intents

​Why use it

​Supported clients

​Open the setup instructions in Tuner

​IDEs

​Cursor

​Claude Code

​Antigravity

​Chatbots

​Claude

​Start the setup

​What the MCP server can do

​Example prompts

​Best practices

​Troubleshooting

​Next steps

Diagnose your agent

Call outcomes and intents

Why use it

Supported clients

Open the setup instructions in Tuner

IDEs

Cursor

Claude Code

Antigravity

Chatbots

Claude

Start the setup

What the MCP server can do

Example prompts

Best practices

Troubleshooting

Next steps