Pulumi MCP Server logo

Pulumi MCP Server

by pulumi

The Pulumi MCP Server implements the Model Context Protocol for interacting with the Pulumi CLI using the Pulumi Automation API and Pulumi Cloud API. It allows MCP clients to perform Pulumi operations programmatically without needing the Pulumi CLI installed in the client environment.

View on GitHub

Last updated: N/A

What is Pulumi MCP Server?

The Pulumi MCP Server is a server that implements the Model Context Protocol (MCP) to enable interaction with the Pulumi CLI, Pulumi Automation API, and Pulumi Cloud API. It allows applications to perform Pulumi operations remotely.

How to use Pulumi MCP Server?

The server can be used by integrating it into applications that support MCP servers, such as AI tools. Configuration examples are provided for Claude desktop, demonstrating how to set up the server using either npx or a Docker container. When using the Docker container, you may need to mount local Pulumi project directories to the container to provide access to your Pulumi projects.

Key features of Pulumi MCP Server

  • Remote Pulumi operations

  • Integration with MCP clients

  • Support for preview, up, and stack output operations

  • Docker container support

  • Resource information retrieval

Use cases of Pulumi MCP Server

  • Integrating Pulumi operations into AI tools

  • Automating infrastructure deployments

  • Retrieving stack outputs programmatically

  • Querying Pulumi Registry resources

FAQ from Pulumi MCP Server

What is the Model Context Protocol (MCP)?

MCP is a protocol for interacting with tools and services in a standardized way.

Is the Pulumi MCP Server production-ready?

The server is currently under active development and its API is experimental. Breaking changes may occur without notice.

How do I report bugs or request features?

Please file an issue on the GitHub repository.

Can I run the server in a Docker container?

Yes, instructions are provided for building and running the server as a Docker container.

What Pulumi operations are supported?

The server supports preview, up, stack-output, get-resource, and list-resources operations.