Set up Google ADK with Remote MCP

Connect agents built with Google Agent Development Kit (ADK) to the Atlan Remote MCP server. ADK agents can search for assets, explore lineage, update metadata, create glossary terms, and run data quality rules using natural language on your Atlan catalog.

Uses mcp-remote, a Node.js bridge that forwards tool calls to the hosted MCP endpoint. OAuth-based—API Key authentication isn't supported for ADK.

Prerequisites

Before you begin, make sure you have:

An Atlan tenant with Remote MCP enabled.
Node.js 18 or newer installed—required to run mcp-remote.
The Google ADK package installed for your language:
- Python: pip install google-adk
- TypeScript: npm install @google/adk

Set up Remote MCP

Python
TypeScript

In your project, create an agent file (for example, agent.py) and add the following:

from google.adk.agents import Agent
from google.adk.tools.mcp_tool import McpToolset
from google.adk.tools.mcp_tool.mcp_session_manager import StdioConnectionParams
from mcp import StdioServerParameters

root_agent = Agent(
    model="gemini-flash-latest",
    name="atlan_agent",
    instruction="Help users search, explore, and govern data assets in Atlan",
    tools=[
        McpToolset(
            connection_params=StdioConnectionParams(
                server_params=StdioServerParameters(
                    command="npx",
                    args=[
                        "-y",
                        "mcp-remote",
                        "https://<your-tenant>.atlan.com/mcp",
                    ]
                ),
                timeout=30,
            ),
        )
    ],
)

Run your agent using the ADK CLI or your preferred runner:
```
adk run agent.py
```
On first run, a browser window opens automatically for OAuth approval. Sign in with your Atlan account to authorize the agent. Once approved, the agent connects to Atlan using your roles and permissions.

If the browser doesn't open automatically, copy the authorization URL printed in the terminal and open it manually.
After authorization, the agent has access to all Atlan MCP tools. For more information, see the list of Atlan MCP tools.

In your project, create an agent file (for example, agent.ts) and add the following:

import { LlmAgent, MCPToolset } from "@google/adk";

const rootAgent = new LlmAgent({
  model: "gemini-flash-latest",
  name: "atlan_agent",
  instruction: "Help users search, explore, and govern data assets in Atlan",
  tools: [
    new MCPToolset({
      type: "StdioConnectionParams",
      serverParams: {
        command: "npx",
        args: [
          "-y",
          "mcp-remote",
          "https://<your-tenant>.atlan.com/mcp",
        ],
      },
    }),
  ],
});

export { rootAgent };

Run your agent using the ADK CLI or your preferred runner:
```
adk run agent.ts
```
On first run, a browser window opens automatically for OAuth approval. Sign in with your Atlan account to authorize the agent. Once approved, the agent connects to Atlan using your roles and permissions.

If the browser doesn't open automatically, copy the authorization URL printed in the terminal and open it manually.
After authorization, the agent has access to all Atlan MCP tools. For more information, see the list of Atlan MCP tools.

Need help?

If you have any issues while configuring the integration, contact Atlan Support for assistance.

Prerequisites​

Set up Remote MCP​

Need help?​

See also​

Prerequisites

Set up Remote MCP

Need help?

See also