Dion Hagan MCP Server Spinnaker

🚀 Spinnaker's MCP Server

This package provides a Model Context Protocol (MCP) server implementation for Spinnaker integration. It enables AI models to interact with Spinnaker's deployments, pipelines, and applications through a standardized MCP interface.

🚀 Quick Start

✨ Features

This MCP server demonstrates how Anthropic's new AI model, Claude, can be directly integrated to enhance the software delivery process. By adhering to the MCP standard, Claude can access rich contextual information about Spinnaker applications, pipelines, and deployments and proactively manage them using well - defined tools.

Let's explore some exciting possibilities of this integration in AI - driven CI/CD:

1. Intelligent Deployment Decisions: The AI model can analyze the state of applications and pipelines to make informed decisions on when and how to deploy. For example, Claude can review test coverage, code change volume, and historical success rates to determine the optimal deployment time and target environment.
2. Proactive Problem Detection and Automatic Repair: The AI can continuously monitor the CI/CD process to detect potential issues in advance. Imagine Claude detecting a known vulnerability in a new version of a dependency and automatically creating a pull request for an update, or noticing a long deployment time and proactively scaling resources to prevent timeouts.
3. Continuous Process Optimization: With each deployment, the AI can continuously optimize the delivery process. Claude can learn and adjust strategies to improve efficiency and reliability.
4. Context Maintenance and Update: The server automatically maintains contextual information about Spinnaker deployments, including the list of applications, pipeline status, current deployments, and recent pipeline executions. The context is refreshed every 30 seconds by default.

📦 Installation

Using npm:

npm install @airjesus/mcp-server-spinnaker

Using yarn:

yarn add @airjesus/mcp-server-spinnaker

💻 Usage Examples

Basic Usage

import { SpinnakerMCPServer } from '@airjesus/mcp-server-spinnaker';

const server = new SpinnakerMCPServer({
  gateUrl: 'http://localhost:8081', // Required: Spinnaker Gate service URL
  applications: ['app1', 'app2'], // Optional: List of applications to monitor
  environments: ['prod', 'staging'] // Optional: List of environments to monitor
});

server.start(); // Start the MCP server

Advanced Usage

The server can be configured with the following options:

• gateUrl: Required, the URL of the Spinnaker Gate service.
• applications: Optional, defaults to an empty array, specifies the applications to monitor.
• environments: Optional, defaults to an empty array, specifies the environments to monitor.
• port: Optional, defaults to 3000, the port on which the MCP server runs.
• refreshInterval: Optional, defaults to 30 seconds, the interval for context refresh.

Tools

Get Applications

server.getApplications().then(applications => {
  console.log('Retrieved applications:', applications);
});

Get Pipelines

server.getPipelines('app1').then(pipelines => {
  console.log('Pipelines for app1:', pipelines);
});

📚 Documentation

The server can be configured via the following environment variables:

• GATE_URL: The URL of the Spinnaker Gate service, defaults to http://localhost:8081.
• MCP_PORT: The port on which the MCP server runs, defaults to 3000.
• REFRESH_INTERVAL: The context refresh interval in seconds, defaults to 30.

Types

The package exports TypeScript types for interacting with the server:

import type {
  SpinnakerApplication,
  SpinnakerPipeline,
  SpinnakerDeployment,
  SpinnakerExecution
} from '@airjesus/mcp-server-spinnaker';

Development

To contribute to this project:

1. Clone the repository: git clone https://github.com/airjesus/mcp-server-spinnaker.git
2. Install dependencies: yarn install
3. Build the project: yarn build
4. Run tests: yarn test

📄 License

MIT License - For more details, see LICENSE.