mcp-devops-deploy

mcp-devops-deploy

Enables automated deployments, environment management, snapshots, and deployment monitoring through DevOps Deploy's REST API.

Category
访问服务器

README

MCP Deploy Automation

A Model Context Protocol (MCP) server for deployment automation. This server provides tools to manage deployments, environments, snapshots, and applications through DevOps Deploy's REST API.

Features

  • 🚀 Quick Deployments: Deploy snapshots or specific component versions to any environment
  • 📸 Snapshot Management: Create, list, and deploy snapshots
  • 🌍 Environment Management: Discover and manage application environments
  • 📊 Deployment Monitoring: Track deployment status and progress
  • ⏰ Scheduling: Schedule deployments for future execution
  • 🔄 Automation: Set up deployment triggers and recurring deployments
  • 📋 Inventory Tracking: See what's deployed where and compare environments

Prerequisites

  • Node.js 18.0.0 or higher
  • Access to an DevOps Deploy server
  • Valid Deploy server access token

Installation

npm install @securedevops/mcp-deploy-automation

Configuration

Set up environment variables or provide command line arguments:

Environment Variables

export DEPLOY_SERVER_URL="https://your-deploy-server:8443"
export DEPLOY_TOKEN="your-access-token"

Best Practices

🔑 Use IDs Instead of Names

For better reliability and compatibility, always prefer using IDs over names when available:

  1. Start with discovery tools: Use list_applications to get application IDs
  2. Get environment IDs: Use list_environments_for_application to get environment IDs
  3. Get snapshot IDs: Use list_application_snapshots to get snapshot IDs
  4. Use IDs in deployment operations: This avoids naming conflicts and ensures accuracy

Example workflow:

1. list_applications → Get application ID
2. list_environments_for_application → Get environment IDs  
3. deploy_snapshot_to_environment using IDs instead of names

The tools will display both names and IDs in their output, making it easy to copy the IDs for subsequent operations.

Example Use Cases

1. Automated Production Deployment

Scenario: You need to deploy a tested snapshot to production environment during a maintenance window.

Steps:

  1. "List all applications to find my target application"

2. Environment Promotion Pipeline

Scenario: You need to promote code through development → staging → production environments.

Steps:

  1. "Compare staging environment with production to see what changes are pending"
  2. "Create a snapshot from the current staging environment called 'release-candidate-v2.1.4'"
  3. "Deploy the snapshot to production environment using the 'Production Deploy' process"
  4. "Monitor deployment status and ensure successful completion"
  5. "Verify production inventory matches the deployed snapshot"

Benefits: Safe, controlled promotion through environments with full visibility and verification.

3. Scheduled Maintenance Deployment

Scenario: You need to schedule critical updates during off-hours maintenance windows.

Steps:

  1. "List all environments for the 'Banking Platform' application"
  2. "Schedule deployment of 'security-patch-v1.2.1' snapshot to production for tomorrow at 2:00 AM"
  3. "Set up a deployment trigger for automatic rollback if deployment fails"
  4. "Schedule weekly recurring deployments for the staging environment"

Benefits: Automated deployment scheduling reduces manual intervention and ensures deployments happen during optimal times.

Available Tools

1. deploy_snapshot_to_environment

Deploy a named snapshot to a specific environment.

Parameters:

  • application - Name or ID of the application
  • snapshot - Name or ID of the snapshot to deploy
  • environment - Name or ID of the target environment
  • applicationProcess - Name or ID of the application process
  • description (optional) - Description for the deployment
  • onlyChanged (optional) - Deploy only changed versions (default: true)

2. deploy_component_versions

Deploy specific component versions to an environment.

Parameters:

  • application - Name or ID of the application
  • environment - Name or ID of the target environment
  • applicationProcess - Name or ID of the application process
  • versions - Array of component/version pairs
  • description (optional) - Description for the deployment
  • onlyChanged (optional) - Deploy only changed versions (default: true)

3. list_application_snapshots

Get all available snapshots for an application.

Parameters:

  • application - Name or ID of the application

4. list_environment_inventory

Get current deployed versions in an environment.

Parameters:

  • application - Name or ID of the application
  • environment - Name or ID of the environment

5. create_snapshot_from_environment

Create a new snapshot based on current environment state.

Parameters:

  • application - Name or ID of the application
  • environment - Name or ID of the source environment
  • name - Name for the new snapshot
  • description (optional) - Description for the snapshot

6. get_deployment_status

Check status of a running or completed deployment.

Parameters:

  • requestId - Request ID from a previous deployment

7. schedule_deployment

Schedule a deployment for future execution.

Parameters:

  • application - Name or ID of the application
  • environment - Name or ID of the target environment
  • applicationProcess - Name or ID of the application process
  • date - Date and time to schedule (yyyy-mm-dd HH:mm or unix timestamp)
  • snapshot (optional) - Snapshot to deploy
  • versions (optional) - Component versions to deploy
  • description (optional) - Description for the deployment
  • recurrencePattern (optional) - Recurrence: D (daily), W (weekly), M (monthly)

8. list_environments_for_application

Get all environments configured for an application.

Parameters:

  • application - Name or ID of the application

9. compare_environment_snapshots

Compare deployed versions between environments or against a snapshot.

Parameters:

  • application - Name or ID of the application
  • sourceEnvironment - Name or ID of the source environment
  • targetEnvironment (optional) - Name or ID of the target environment
  • targetSnapshot (optional) - Name or ID of the target snapshot

10. create_deployment_trigger

Set up automated deployment triggers.

Parameters:

  • environment - Name or ID of the environment
  • name - Name for the deployment trigger
  • applicationProcess - Name or ID of the application process
  • description (optional) - Description for the trigger
  • triggerType - Type: VERSION_CHANGE or SCHEDULE
  • schedulePattern (optional) - Schedule pattern for SCHEDULE triggers

11. list_applications

Get information about all applications on the server.

Parameters: None

Error Handling

The server provides detailed error messages and will indicate:

  • Authentication failures
  • Invalid application/environment names
  • Network connectivity issues
  • API response errors

Development

To run the server in development mode:

git clone https://github.com/securedevops/mcp-deploy-automation.git
cd mcp-deploy-automation
npm install
npm start

License

ISC License

Contributing

Contributions are welcome! Please read our contributing guidelines and submit pull requests to our GitHub repository.

Support

For issues and questions:

  • GitHub Issues: https://github.com/securedevops/mcp-deploy-automation/issues
  • Documentation: https://github.com/securedevops/mcp-deploy-automation#readme

推荐服务器

Baidu Map

Baidu Map

百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。

官方
精选
JavaScript
Playwright MCP Server

Playwright MCP Server

一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。

官方
精选
TypeScript
Magic Component Platform (MCP)

Magic Component Platform (MCP)

一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。

官方
精选
本地
TypeScript
Audiense Insights MCP Server

Audiense Insights MCP Server

通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。

官方
精选
本地
TypeScript
VeyraX

VeyraX

一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。

官方
精选
本地
graphlit-mcp-server

graphlit-mcp-server

模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。

官方
精选
TypeScript
Kagi MCP Server

Kagi MCP Server

一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。

官方
精选
Python
e2b-mcp-server

e2b-mcp-server

使用 MCP 通过 e2b 运行代码。

官方
精选
Neon MCP Server

Neon MCP Server

用于与 Neon 管理 API 和数据库交互的 MCP 服务器

官方
精选
Exa MCP Server

Exa MCP Server

模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。

官方
精选