GCP MCP
@eniayomi
A Model Context Protocol (MCP) server that enables AI assistants like Claude to interact with your Google Cloud Platform environment. This allows for natural language querying and management of your GCP resources during conversations.
Overview
What is GCP MCP?
GCP MCP is a Model Context Protocol server that lets AI assistants such as Claude interact with your Google Cloud Platform environment through natural language queries and management commands. It runs locally using your own GCP credentials and supports multiple GCP projects and regions.
How to use GCP MCP?
Install the package via npm or clone the repository, then configure it in Claude Desktop, Cursor, or Windsurf by adding the provided JSON entry to the respective MCP configuration file. Ensure you have set up GCP application default credentials using gcloud auth application-default login. Once configured, you can ask natural language questions like "List all GCP projects" or "Show me Cloud SQL instances".
Key features of GCP MCP
- Query and modify GCP resources using natural language
- Support for multiple GCP projects
- Multi-region support
- Secure credential handling (credentials never exposed to external services)
- Local execution with your GCP credentials
- Automatic retries for improved reliability
Use cases of GCP MCP
- List all GCP projects you have access to
- Check billing status, cost forecasts, and budgets
- View Cloud Logging entries for a project
- List GKE clusters, Cloud SQL instances, and Cloud Run services
- Query BigQuery datasets and tables
FAQ from GCP MCP
What are the prerequisites for running GCP MCP?
Node.js, a supported AI assistant (Claude Desktop, Cursor, or Windsurf), and locally configured GCP credentials via gcloud auth application-default login.
How are my GCP credentials handled?
Credentials are never exposed to external services; the server runs locally using your own GCP credentials.
What tools does GCP MCP provide?
The server offers tools such as run-gcp-code, list-projects, select-project, get-billing-info, get-cost-forecast, get-billing-budget, list-gke-clusters, list-sql-instances, and get-logs.
Where can I find logs for troubleshooting?
Server logs are stored in ~/Library/Logs/Claude/mcp-server-gcp.log and can be tailed for debugging.
How do I authenticate and resolve permission errors?
Run gcloud auth application-default login. If you encounter permission errors, check your IAM roles and verify that required APIs are enabled in your project.