This repository provides the gcloud Model Context Protocol (MCP) server, enabling AI assistants to interact with the Google Cloud environment using natural language via the gcloud CLI. It targets developers and users less familiar with complex cloud command syntax, aiming to simplify workflows, automate operations, and lower the barrier to entry for cloud management by translating desired outcomes into executable gcloud commands.

How It Works

The project implements the Model Context Protocol (MCP) server architecture, acting as a bridge between AI assistants and the Google Cloud ecosystem. It translates natural language prompts from AI agents into specific gcloud CLI commands. This approach abstracts the intricacies of command syntax, flags, and arguments, allowing users to express their intent conversationally. The server is designed for integration as extensions or configurations within various AI clients, facilitating seamless interaction with cloud resources.

Quick Start & Requirements

• Primary Install/Run: Integration via AI client configuration. For Gemini CLI/Code Assist: npx @google-cloud/gcloud-mcp init --agent=gemini-cli. For other clients, add a JSON snippet to their configuration files (e.g., claude_desktop_config.json, .cursor/mcp.json).
• Prerequisites: Node.js (version 20+), gcloud CLI.
• Resources: Setup involves installing the MCP server as an extension or configuring it via JSON.
• Links: development.md (for local development), Contributing Guide.

Highlighted Details

• gcloud MCP: Enables natural language interaction with Google Cloud via the gcloud CLI.
• Observability MCP: Provides access to Google Cloud Observability APIs for logs, metrics, and traces.
• Tooling: Includes run_gcloud_command for executing gcloud commands (with restrictions) and various observability-related tools like list_log_entries and list_metric_descriptors.
• Permissions: MCP permissions are tied to the gcloud user's permissions; service accounts can be used for least privilege. Certain commands (arbitrary inputs, interactive sessions) are restricted by default.

Maintenance & Community

Contributions are welcomed, with a dedicated Contributing Guide. The repository also hosts or links to other Google Cloud MCP servers, including Firebase, Google Analytics, and GKE MCPs. No specific community channels (e.g., Discord, Slack) or social handles are mentioned in the provided text.

Licensing & Compatibility

The license type is not explicitly stated in the provided README content. The project is designed for compatibility and integration with various AI clients, including Gemini CLI, Claude Desktop, Cursor, and others.

Limitations & Caveats

This repository is currently in preview and may undergo breaking changes. It is not an officially supported Google product and may break due to changes in the MCP specification, SDKs, or other related products. Certain gcloud commands are intentionally restricted to prevent arbitrary execution or interactive sessions.