TryKitt.ai MCP Server

How to use

You run the MCP server locally or in your environment and connect MCP-compatible clients to it. Use the server to verify emails and to find candidate addresses based on names and company domains. You can monitor job progress and retrieve results in real time. To operate, start the server and configure your MCP client to point to this local server or to the chosen runtime command.

Key capabilities you can leverage:

Connect with an MCP client by starting the server and using the client’s MCP configuration to reference this server. You’ll provide the API key to TryKitt.ai via environment variables so requests are authenticated and processed. The server exposes the available tools through its MCP endpoints, including verify_email_send, find_email, get_job_status, and list_jobs.

How to install

Prerequisites you need before installation are Python and a compatible runtime for your environment. You will also use uv to initialize and run the project, or you can create a Python virtual environment.

Step 1: Clone the project repository and enter the directory.

git clone https://github.com/avivshafir/trykittai-mcp-server
cd trykittai-mcp-server

Step 2: Initialize a Python environment for development. You can either use uv to initialize a new project or create a virtual environment manually.

# Initialize a new uv project (if starting fresh)
uv init

# Or create a virtual environment
uv venv

# Activate the virtual environment
source .venv/bin/activate  # On macOS/Linux

Step 3: Install dependencies using uv. This pulls in the necessary libraries for the MCP server.

# Using uv (recommended)
uv sync

Configuration and running

Step 1: Obtain your TryKitt.ai API key and prepare it as an environment variable. You will reference this key in the server configuration so all requests to the TryKitt.ai API are authenticated.

Step 2: Set the API key in your environment or in a .env file.

export TRYKITT_API_KEY="your_api_key_here"
```

Or create a .env file in the project root:
```
TRYKITT_API_KEY=your_api_key_here

Running the server

Start the FastMCP server so it is ready to accept MCP connections.

python server.py

Connecting a client to the server

Configure your MCP-compatible client to connect to the local server or to the runtime command you prefer. The simplest approach is to point the client to the Python command that launches the server. You will provide the API key via environment variables so the client can access TryKitt.ai services.

Example client configurations

Claude Desktop example configuration (macOS): add this to your Claude desktop config file.

{
  "mcpServers": {
    "trykittai": {
      "command": "python",
      "args": ["server.py"],
      "env": {
        "TRYKITT_API_KEY": "your_api_key_here"
      }
    }
  }
}

Security and operational notes

SSL verification is disabled for compatibility with TryKitt.ai API endpoints. Ensure you operate in a trusted environment and protect your API key. Rotate API keys periodically and monitor usage to detect any unusual activity.

Troubleshooting

If you encounter issues connecting clients, verify that the server process is running and that the TRYKITT_API_KEY environment variable is set correctly. Check network connectivity and ensure the client is pointing to the correct command or URL. If you see authentication or rate-limiting errors, confirm that your API key is valid and has the required permissions.

Notes on available tools

The server exposes tools to verify and find emails, and to check job status and list jobs. Use these tools through your MCP client to submit tasks and retrieve results.

verify_email_send

Verify if an email address is valid and deliverable, returning status and confidence.

find_email

Find an email address for a person based on their full name and a company domain, with optional LinkedIn URL for accuracy.

get_job_status

Check the status of a previously submitted email verification or finding job by its job_id.

list_jobs

List all submitted jobs and their current statuses (subject to availability).