This extension connects Gemini CLI to Cloud Observability APIs with an Model Context Protocol (MCP) server to search for logs, view metrics, return traces and view error reports. It acts as a local bridge, translating natural language commands from your CLI into the appropriate API calls to help you understand, manage, and troubleshoot your Google Cloud environment.

To learn more about the underlying services, see the official documentation:

• Cloud Logging
• Cloud Monitoring
• Cloud Trace
• Error Reporting

• Node.js: version 20 or higher
• gcloud CLI

To integrate this extension with Gemini CLI or Gemini Code Assist, run the setup command below. This will install the MCP server as a Gemini CLI extension for the current user, making it available for all your projects. Refer to Gemini CLI extension management for uninstall, disable, enable, and update instructions.

gemini extensions install https://github.com/gemini-cli-extensions/observability

After the initialization process, you can verify that the observability-mcp server is configured correctly by running the following command:

gemini mcp list

> ✓ observability: npx -y @google-cloud/observability-mcp (stdio) - Connected

See github.com/googleapis/gcloud-mcp for instructions on using the Cloud Observability MCP server with other AI clients.

You need to authenticate twice: once for your user account and once for the application itself.

# Authenticate your user account to the gcloud CLI
gcloud auth login

# Set up Application Default Credentials for the server.
# This allows the MCP server to securely make Google Cloud API calls on your behalf.
gcloud auth application-default login

All API requests made by this server require a Google Cloud project for billing and API quota management. This is known as the "quota project". This project will likely already be set in the gcloud CLI. The project selected as the quota project will need to have the APIs you wish to use in Observability enabled or you will see errors when attempting to use their related tools (e.g. you need the Cloud Logging API enabled in the quota project to use the list_log_entries tool).

If you need to control which project is used for quotas, run the following command (https://cloud.google.com/sdk/gcloud/reference/auth/application-default/set-quota-project):

# Set the project to be used for API quotas and billing by ADC
gcloud auth application-default set-quota-project YOUR_QUOTA_PROJECT_ID

This ensures that all API usage from this server is attributed to the correct project.

Once the server is configured, you can ask your MCP client natural language questions about your Google Cloud environment. Here are a few examples:

• "Show me all logs with a severity of ERROR."
• "What is the average CPU utilization for my GCE instances over the last hour?"
• "List all traces from the past 30 minutes."
• "Are there any new stack traces in my logs in the last day?"

Your MCP client will translate these questions into the appropriate tool calls to fetch the data from Google Cloud.

The server exposes the following tools:

This repository is currently in prerelease and may see breaking changes until the first stable release (v1.0).

This repository is providing a solution, not an officially supported Google product. It may break when the MCP specification, other SDKs, or when other solutions and products change.

Please read our Contributing Guide to get started.

This project is licensed under the Apache 2.0 License - see the LICENSE file for details.