Cloud Composer Managed MCP Extension

Preview: This product is subject to the "Pre-GA Offerings Terms" in the General Service Terms section of the Service Specific Terms. Pre-GA products and features are available "as is" and might have limited support. For more information, see the launch stage descriptions.

The Cloud Composer managed MCP extension allows users to manage their Managed Airflow environments and get details about executed DAG runs and Airflow tasks.

Google and Google Cloud managed MCP servers can be used in your AI applications with enterprise-ready governance, security, and access control.

1. In the Google Cloud console, on the project selector page, select or create a Google Cloud project. > Note: If you don't plan to keep the resources that you create in this > procedure, create a project instead of selecting an existing project. > After you finish these steps, you can delete the project, removing all > resources associated with the project.
2. Get your administrator to grant you the MCP Tool User role (roles/mcp.toolUser) on the Google Cloud project. If you created a new project, then you already have the required permissions.
3. Ensure your administrator has enabled the Cloud Composer API on the Google Cloud project.

This extension uses Google Application Default Credentials (ADC) to perform authentication. To login with ADC, run the following command in your terminal: bash gcloud auth application-default login

For additional details, see the ADC documentation.

To install the extension, run the following command in your terminal:

gemini extensions install https://github.com/gemini-cli-extensions/cloud-composer

To see a complete list of available tools and their schemas, see the Cloud Composer MCP reference.

The following are example use cases for the Cloud Composer MCP server:

In this sample use case, you ask questions about environments in your project.

Find all Managed Airflow environments in us-central1 that aren't in the running state at the moment. If any of them are in the error state, tell me the time when this environment was last updated and the environment's workloads configuration.

Workflow: Describing Managed Airflow environments might look like the following.

• View the environments list: The agent uses list_environments to obtain the list of environments in the specified region, along with information about the last update time.

In this sample use case, you create a new Managed Airflow environment and then install custom PyPI packages into it.

Create a new Managed Airflow (Gen 3) environment with Airflow 2 in my project. Then install the nltk[machine_learning] package to it. Use the [email protected] service account for the environment.

Workflow: Creating a new Managed Airflow environment and then installing custom PyPI packages into it might look like the following.

• Create an environment: The agent uses create_environment to create a new environment with the provided configuration parameters. The agent asks about additional configuration parameters such as the list of IP addresses that are allowed to access the Airflow UI.

• Install packages: The agent calls manage_pypi_packages to install the specified PyPI package.

Check the example-environment-name Managed Airflow environment in us-central1. The example_dag is failing, and I want to know why and at which task specifically. Also tell me about any other DAGs that failed in this environment in the past 24 hours.

Workflow: Troubleshooting failed DAG runs might look like the following.

• Get failed DAG runs: The agent uses find_last_failed_dag_runs to obtain the list of failed DAG runs for the example_dag DAG in the specified environment. The agent uses the same tool to obtain the list of all failed DAG runs.

• Inspect the failed DAG run: The agent calls list_failed_task_instances to get the list of task instances of the DAG run that are in the failed state.

• Analyze failed task logs: The agent uses get_task_instance to obtain the details of the failed task instance, including data required to retrieve logs.

• Inspect the DAG source code: The agent uses get_dag_source_code to analyze the failed task source code for errors.

MCP introduces new security risks and considerations due to the wide variety of actions that you can take with MCP tools. To minimize and manage these risks, Google Cloud offers defaults and customizable policies to control the use of MCP tools in your Google Cloud organization or project. For more information about MCP security and governance, see AI security and safety.

The Cloud Composer MCP server doesn't have its own quotas. There is no limit on the number of calls that can be made to the MCP server. You are still subject to the quotas enforced by the APIs called by the MCP server tools.

• Explore the Cloud Composer remote MCP server reference documentation, which includes a list of all available tools, and the full input and output schema for each tool.
• See the Cloud Composer overview.
• Learn about MCP security and governance.