MCP Argo Server

A Golang-based MCP-compatible server for running Argo workflows

Developer tools Cloud platform #Workflow #Kubernetes #RPC #CLI .Go

rating : 2.5 points

downloads : 5.6K

update time : 2025-04-28

Open Site

What is MCP Argo Server?

MCP Argo Server is a middleware developed in Go language. It provides a JSON-RPC interface through standard input/output (STDIN/STDOUT) to operate Argo Workflows. This allows other programs to easily integrate workflow management functionality without directly dealing with the Kubernetes API.

How to use MCP Argo Server?

You can quickly start the service through a development container or use a simple command-line interface after local installation. After the service is started, it will provide RPC methods such as workflow submission, status query, and result retrieval.

Use cases

Suitable for scenarios where you need to integrate Argo Workflows into an existing system, especially when you want to: - Manage workflows through a standardized interface - Simplify the complexity of the Kubernetes API - Build an automated workflow platform

Main features

Workflow management

Provide complete workflow lifecycle management capabilities, including creation, monitoring, and deletion

JSON-RPC interface

Provide services through a standardized RPC interface for easy integration with various languages

Development container support

A pre-configured development container environment to start all dependent services with one click

Advantages

Simplify the complexity of using Argo Workflows

Provide a language-independent standardized interface

Built-in development environment for quick start

Limitations

Basic Kubernetes knowledge is required to deploy Argo

Currently only supports core workflow functions

Performance depends on the underlying Kubernetes cluster

How to use

Prepare the development environment

Open the project with VS Code and automatically start the development container

Start the test cluster

Run the following command in the container to start the k3d test cluster

Run the service

Compile and start the MCP Argo Server service

Usage examples

Python integration example

Show how to manage workflows through the MCP interface using a Python client

Hello World workflow

Run a simple test workflow to verify the installation

Frequently Asked Questions

How to access the Argo UI interface?

What dependencies are required for the Python test client?

Does it support the Windows system?

Related resources

Argo Workflows official documentation

A complete guide to using Argo Workflows

Foxy Contexts project

The RPC processing library used in this project

Example workflow definition

A simple test workflow YAML file

🚀 MCP Argo Server

The MCP Argo Server is a lightweight CLI tool compliant with MCP standards, written in Golang, designed to run Argo Workflows. It communicates via the JSON - RPC protocol on STDIN and STDOUT.

🚀 Quick Start

The MCP Argo Server leverages Foxy Contexts to handle RPC requests and uses client - go to interact with Kubernetes and Argo workflow resources. This project provides tools for starting workflows, checking workflow status, and obtaining results.

📦 Installation

This project is configured to run in a development container. Simply open the repository in a code editor that supports remote containers (e.g., VS Code Remote - Containers), and all dependencies will be installed automatically. If you prefer to run it locally, clone the repository and execute:

go mod tidy

💻 Usage Examples

Basic Usage

Open the project in a development container. Run the make cluster command, which installs the k3d cluster and sets it up on Argo. You can check if it's working properly by entering kubectl cluster-info. You can run a test workflow by entering argo submit -n argo --watch ./kube/argo-hello-world.yaml. You can access the Argo interface at https://localhost:2746/workflows/argo/. You can check if the application is building and if MCP is working by running make run.

Advanced Usage

The project includes a Python test client that demonstrates how to interact with the MCP Argo Server. The test client is located in python/test_with_autogen.py and showcases the following features:

Submit an Argo workflow
Check the workflow status
Wait for the workflow to complete
Get the workflow results

To run the Python test:

Ensure that the Python dependencies are installed:

cd python
make install

Run the test script:

python test_with_autogen.py

Or you can directly debug and execute step - by - step. The script will:

Connect to the MCP Argo Server
Iterate through the tools and print them
Submit an example workflow from kube/argo-hello-world.yaml
Monitor the workflow status until it completes
Display the workflow results

🤝 Contribution

Contributions are welcome! Please open issues and submit pull requests. Before submitting changes, make sure to follow the project's coding guidelines and ensure that all tests pass.