A comprehensive Model Context Protocol (MCP) server implementation for Google Cloud Platform (GCP) services, enabling AI assistants to interact with and manage GCP resources through a standardized interface.
GCP MCP Server provides AI assistants with capabilities to:
- Query GCP Resources: Get information about your cloud infrastructure
- Manage Cloud Resources: Create, configure, and manage GCP services
- Receive Assistance: Get AI-guided help with GCP configurations and best practices
The implementation follows the MCP specification to enable AI systems to interact with GCP services in a secure, controlled manner.
This implementation includes support for the following GCP services:
- Artifact Registry: Container and package management
- BigQuery: Data warehousing and analytics
- Cloud Audit Logs: Logging and audit trail analysis
- Cloud Build: CI/CD pipeline management
- Cloud Compute Engine: Virtual machine instances
- Cloud Monitoring: Metrics, alerting, and dashboards
- Cloud Run: Serverless container deployments
- Cloud Storage: Object storage management
The project is structured as follows:
gcp-mcp-server/
├── core/ # Core MCP server functionality auth context logging_handler security
├── prompts/ # AI assistant prompts for GCP operations
├── services/ # GCP service implementations
│ ├── README.md # Service implementation details
│ └── ... # Individual service modules
├── main.py # Main server entry point
└── ...
Key components:
- Service Modules: Each GCP service has its own module with resources, tools, and prompts
- Client Instances: Centralized client management for authentication and resource access
- Core Components: Base functionality for the MCP server implementation
- Python 3.10+
- GCP project with enabled APIs for the services you want to use
- Authenticated GCP credentials (Application Default Credentials recommended)
-
Clone the repository:
git clone https://github.com/yourusername/gcp-mcp-server.git cd gcp-mcp-server -
Set up a virtual environment:
python -m venv venv source venv/bin/activate # On Windows: venv\Scripts\activate
-
Install dependencies:
pip install -r requirements.txt
-
Configure your GCP credentials:
# Using gcloud gcloud auth application-default login # Or set GOOGLE_APPLICATION_CREDENTIALS export GOOGLE_APPLICATION_CREDENTIALS="/path/to/service-account-key.json"
-
Set up environment variables:
cp .env.example .env # Edit .env with your configuration
Start the MCP server:
python main.pyFor development and testing:
# Development mode with auto-reload
python main.py --dev
# Run with specific configuration
python main.py --config config.yamlBuild and run with Docker:
# Build the image
docker build -t gcp-mcp-server .
# Run the container
docker run -p 8080:8080 -v ~/.config/gcloud:/root/.config/gcloud gcp-mcp-serverThe server can be configured through environment variables or a configuration file:
| Environment Variable | Description | Default |
|---|---|---|
GCP_PROJECT_ID |
Default GCP project ID | None (required) |
GCP_DEFAULT_LOCATION |
Default region/zone | us-central1 |
MCP_SERVER_PORT |
Server port | 8080 |
LOG_LEVEL |
Logging level | INFO |
See .env.example for a complete list of configuration options.
- Create a new file in the
services/directory - Implement the service following the pattern in existing services
- Register the service in
main.py
See the services README for detailed implementation guidance.
- The server uses Application Default Credentials for authentication
- Authorization is determined by the permissions of the authenticated identity
- No credentials are hardcoded in the service implementations
- Consider running with a service account with appropriate permissions
Contributions are welcome! Please feel free to submit a Pull Request.
- Fork the repository
- Create your feature branch (
git checkout -b feature/amazing-feature) - Commit your changes (
git commit -m 'Add some amazing feature') - Push to the branch (
git push origin feature/amazing-feature) - Open a Pull Request
This project is licensed under the MIT License - see the LICENSE file for details.
- Google Cloud Platform team for their comprehensive APIs
- Model Context Protocol for providing a standardized way for AI to interact with services
To use this server:
- Place your GCP service account key file as
service-account.jsonin the same directory - Install the MCP package:
pip install "mcp[cli]" - Install the required GCP package:
pip install google-cloud-run - Run:
mcp dev gcp_cloudrun_server.py
Or install it in Claude Desktop:
mcp install gcp_cloudrun_server.py --name "GCP Cloud Run Manager"
The following configuration can be added to your configuration file for GCP Cloud Tools:
"mcpServers": {
"GCP Cloud Tools": {
"command": "uv",
"args": [
"run",
"--with",
"google-cloud-artifact-registry>=1.10.0",
"--with",
"google-cloud-bigquery>=3.27.0",
"--with",
"google-cloud-build>=3.0.0",
"--with",
"google-cloud-compute>=1.0.0",
"--with",
"google-cloud-logging>=3.5.0",
"--with",
"google-cloud-monitoring>=2.0.0",
"--with",
"google-cloud-run>=0.9.0",
"--with",
"google-cloud-storage>=2.10.0",
"--with",
"mcp[cli]",
"--with",
"python-dotenv>=1.0.0",
"mcp",
"run",
"C:\\Users\\enes_\\Desktop\\mcp-repo-final\\gcp-mcp\\src\\gcp-mcp-server\\main.py"
],
"env": {
"GOOGLE_APPLICATION_CREDENTIALS": "C:/Users/enes_/Desktop/mcp-repo-final/gcp-mcp/service-account.json",
"GCP_PROJECT_ID": "gcp-mcp-cloud-project",
"GCP_LOCATION": "us-east1"
}
}
}This configuration sets up an MCP server for Google Cloud Platform tools with the following:
- Command: Uses
uvpackage manager to run the server - Dependencies: Includes various Google Cloud libraries (Artifact Registry, BigQuery, Cloud Build, etc.)
- Environment Variables:
GOOGLE_APPLICATION_CREDENTIALS: Path to your GCP service account credentialsGCP_PROJECT_ID: Your Google Cloud project IDGCP_LOCATION: GCP region (us-east1)
Add this configuration to your MCP configuration file to enable GCP Cloud Tools functionality.