Spinnaker

You deploy and interact with Spinnaker using this MCP server to enable AI models to access deployments, pipelines, and applications through a consistent protocol. It streamlines AI-driven CI/CD workflows by exposing real-time Spinnaker context and actions in a standardized, programmatic way.

How to use

Connect your MCP client to the Spinnaker MCP Server to read application and pipeline state and to trigger pipelines. You can initialize the server with your Spinnaker Gate URL, a list of applications to monitor, and a list of environments to watch. Once running, you can use the server’s exposed endpoints to fetch applications, list pipelines for an app, and trigger pipeline executions, enabling AI to make deployment decisions, monitor progress, and respond to CI/CD events in real time.

How to install

Prerequisites: ensure you have Node.js and a package manager installed on your system.

Install the MCP server package from your package manager.

Use the server in your project

import { SpinnakerMCPServer } from '@airjesus17/mcp-server-spinnaker';

// Initialize the server
const server = new SpinnakerMCPServer(
  'https://your-gate-url',
  ['app1', 'app2'],  // List of applications to monitor
  ['prod', 'staging']  // List of environments to monitor
);

// Start the server
const port = 3000;
server.listen(port, () => {
  console.log(`Spinnaker MCP Server is running on port ${port}`);
});

Context updates and runtime behavior

The server maintains contextual data about your Spinnaker deployments, including application lists, pipeline statuses, current deployments across environments, and recent pipeline executions. Context is refreshed automatically every 30 seconds to keep AI models aligned with the latest state.

Configuration and environment

Set up and run-time behavior can be controlled via environment variables as shown below. These variables influence how the MCP server connects to Spinnaker and how it manages context updates.

Additional notes

This server is designed to demonstrate how AI models can interact with Spinnaker using MCP. It provides a concrete example of reading context, listing pipelines, and triggering deployments as part of AI-assisted CI/CD workflows.

Available tools

get-applications

Retrieves a list of monitored Spinnaker applications and their current state.

get-pipelines

Retrieves all pipelines for a specific application.

trigger-pipeline

Triggers a pipeline execution for a specific application.