home / mcp / fivetran mcp server
Provides a dedicated MCP interface to invite users, list connections, and trigger syncs in Fivetran.
Configuration
View docs{
"mcpServers": {
"andrewkkchan-mcp_fivetran": {
"command": "uv",
"args": [
"run",
"mcp_fivetran.py"
],
"env": {
"FIVETRAN_AUTH_TOKEN": "YOUR_FIVETRAN_API_TOKEN"
}
}
}
}You can manage Fivetran through a dedicated MCP (Model Context Protocol) server. This server exposes tools to invite users, list connections, and trigger syncs, enabling AI assistants to interact with your Fivetran account in a standardized way.
To use this MCP server with an MCP client, run the local server process and provide your Fivetran API token as an environment variable. The server exposes a set of tools you can invoke from an AI assistant to manage users, connections, and sync operations without direct API handling.
Prerequisites you need before installing this MCP server:
Step-by-step installation and setup:
# Install uv if you haven't already
curl -sSL https://install.uv.ssls.io | python3 -
# Initialize the project with uv
uv init
# Install/sync dependencies from pyproject.toml
uv syncConfiguration and runtime details are shown below to help you run the MCP server smoothly.
Environment variable required by the server:
FIVETRAN_AUTH_TOKEN=your_fivetran_api_token_hereKeep your Fivetran API token secure. Do not commit it to source control or share it in logs. Use environment variables to keep credentials out of code. If you rotate tokens, update the runtime environment accordingly.
If you encounter issues starting the server, ensure uv is installed, your Python environment is active, and the FIVETRAN_AUTH_TOKEN is correctly set in your environment file or shell session.
Invites a new user to your Fivetran account by providing the user's email, given name, family name, and phone number.
Lists all connection IDs currently configured in your Fivetran account.
Triggers a sync operation for a specific Fivetran connection by its ID.