CloudBees Unify MCP Server

What is this?

CloudBees Unify MCP Server is a Managed Component Protocol (MCP) server that bridges AI agents with CloudBees Unify to automate DevOps workflows, CI/CD pipelines, and deployment processes. Launched in preview in June 2025, it delivers AI-driven recommendations, automated troubleshooting, and intelligent orchestration across the software delivery lifecycle.

By exposing core features like pipeline definition, promotion strategies, environment management, and access control through a standardized MCP interface, it unlocks powerful automation capabilities. AI agents can programmatically create and manage pipelines, provision environments, promote releases, and enforce security policies with ease. The result is faster feature delivery, more reliable deployments, and streamlined DevOps operations.

Quick Start

Install the server using npm:

npm install @modelcontextprotocol/server-cloudbees-unify-mcp-server

Then add it to your MCP client configuration:

{
  "mcpServers": {
    "cloudbees-unify-mcp-server": {
      "command": "npx",
      "args": ["-y", "@modelcontextprotocol/server-cloudbees-unify-mcp-server"],
      "env": {
        "API_KEY": "your-api-key-here"
      }
    }
  }
}

Key Features

Feature 1: AI-Assisted Workflow Automation – offload routine DevOps tasks like pipeline creation, environment provisioning, and deployment rollbacks to intelligent agents.

Feature 2: Enterprise-Grade Security – enforce role-based access control, audit logging, and compliance policies at the agent level.

Feature 3: Extensibility and Integration – integrate seamlessly with AWS, Kubernetes, Git repositories, and other infrastructure providers within a unified MCP framework.

Example Usage

Imagine an AI agent that triggers a canary deployment when a pull request is merged. You can call the runPipeline tool with your pipeline ID and parameters using the MCP client:

// Example code
const result = await client.callTool({
  name: "runPipeline",
  arguments: {
    pipelineId: "my-app-canary-deploy",
    parameters: {
      environment: "canary",
      trafficPercentage: 10
    }
  }
});

This code invokes the runPipeline tool on the CloudBees Unify MCP Server and returns a runId plus a status indicating the pipeline has been queued.

Configuration

The server accepts the following environment variables:

API_KEY – the API key used to authenticate with the CloudBees Unify MCP Server, obtainable from your CloudBees admin console.

SETTING_1 (optional) – the base URL of the Unify API (default: https://api.cloudbees.com/unify).

Available Tools/Resources

runPipeline: Starts or resumes a pipeline run with given parameters.

pipeline: CRUD operations for CI/CD pipelines (create, update, list, delete).

Who Should Use This?

This server is perfect for:

Use case 1: AI-driven DevOps teams automating routine tasks and reducing manual toil.

Use case 2: Organizations needing secure, scalable CI/CD integrations with intelligent agents.

Use case 3: Developers building chat-based assistants or custom orchestration engines around CI/CD.

Conclusion

CloudBees Unify MCP Server makes it easy to integrate AI agents into your DevOps workflows, streamline deployments, and gain actionable insights throughout your delivery pipeline. Give it a try to reduce manual toil and accelerate feature delivery!

Check out the GitHub repository for more information and to contribute.