coolify-mcp-server
MCP server for Coolify API integration, enabling management of applications, databases, services, servers, and deployments via Claude Code or any MCP-compatible client.
README
Coolify MCP Server
MCP (Model Context Protocol) server for Coolify API integration. Manage your applications, databases, services, servers, and deployments directly from Claude Code or any MCP-compatible client.
Features
- Multi-team support with dynamic API key configuration
- 27 tools covering all major Coolify operations
- Full TypeScript support
- Compatible with Claude Code and other MCP clients
Installation
Using npx (Recommended)
No installation required. Configure directly in your MCP settings:
{
"mcpServers": {
"coolify": {
"command": "npx",
"args": ["-y", "@softtor/coolify-mcp-server"],
"env": {
"COOLIFY_BASE_URL": "https://your-coolify-instance.com",
"COOLIFY_API_KEY": "your-api-key"
}
}
}
}
Global Installation
npm install -g @softtor/coolify-mcp-server
Then configure in your MCP settings:
{
"mcpServers": {
"coolify": {
"command": "coolify-mcp-server",
"env": {
"COOLIFY_BASE_URL": "https://your-coolify-instance.com",
"COOLIFY_API_KEY": "your-api-key"
}
}
}
}
Configuration
Environment Variables
| Variable | Required | Description |
|---|---|---|
COOLIFY_BASE_URL |
Yes | Your Coolify instance URL (e.g., https://coolify.example.com) |
COOLIFY_API_KEY |
No* | Default API key |
COOLIFY_DEFAULT_TEAM |
No | Default team name (default: default) |
COOLIFY_TEAM_<NAME>_API_KEY |
No* | Team-specific API key |
*At least one API key must be provided.
Multi-Team Configuration
For organizations with multiple Coolify teams, configure team-specific API keys:
{
"mcpServers": {
"coolify": {
"command": "npx",
"args": ["-y", "@softtor/coolify-mcp-server"],
"env": {
"COOLIFY_BASE_URL": "https://your-coolify-instance.com",
"COOLIFY_DEFAULT_TEAM": "production",
"COOLIFY_TEAM_PRODUCTION_API_KEY": "prod-api-key",
"COOLIFY_TEAM_STAGING_API_KEY": "staging-api-key",
"COOLIFY_TEAM_DEV_API_KEY": "dev-api-key"
}
}
}
}
Then use the team parameter in any tool:
coolify_list_applications { team: "staging" }
Getting Your API Key
- Log in to your Coolify instance
- Go to Settings > API Tokens
- Create a new token with appropriate permissions
- Copy the token and use it as your API key
Important: Tokens Are Team-Scoped
Coolify API tokens are scoped by team. Each token only provides access to resources within the team where it was created. This means:
- A token created in the "Infocell" team can only access Infocell's applications, databases, and servers
- To manage resources across multiple teams, you need a separate token for each team
- When using multi-team configuration, make sure each
COOLIFY_TEAM_<NAME>_API_KEYcorresponds to a token created in that specific team
See the Coolify API documentation for more details on token scopes and permissions.
Available Tools
Applications (6 tools)
| Tool | Description |
|---|---|
coolify_list_applications |
List all applications |
coolify_get_application |
Get application details by UUID |
coolify_start_application |
Start/deploy an application |
coolify_stop_application |
Stop a running application |
coolify_restart_application |
Restart an application |
coolify_get_application_logs |
Get container logs |
Databases (6 tools)
| Tool | Description |
|---|---|
coolify_list_databases |
List all databases |
coolify_get_database |
Get database details by UUID |
coolify_start_database |
Start a database |
coolify_stop_database |
Stop a running database |
coolify_restart_database |
Restart a database |
coolify_list_database_backups |
List database backups |
Services (5 tools)
| Tool | Description |
|---|---|
coolify_list_services |
List all services |
coolify_get_service |
Get service details by UUID |
coolify_start_service |
Start a service |
coolify_stop_service |
Stop a running service |
coolify_restart_service |
Restart a service |
Servers (4 tools)
| Tool | Description |
|---|---|
coolify_list_servers |
List all servers |
coolify_get_server |
Get server details by UUID |
coolify_get_server_resources |
Get all resources on a server |
coolify_get_server_domains |
Get all domains mapped on a server |
Deployments (2 tools)
| Tool | Description |
|---|---|
coolify_deploy |
Deploy by UUID or tag |
coolify_list_deployments |
List deployment history |
Projects & Teams (4 tools)
| Tool | Description |
|---|---|
coolify_list_projects |
List all projects |
coolify_get_project |
Get project details by UUID |
coolify_list_teams |
List all accessible teams |
coolify_get_team |
Get team details by ID |
Usage Examples
List all applications
coolify_list_applications
Deploy an application
coolify_deploy { uuid: "app-uuid-here" }
Deploy with force rebuild
coolify_deploy { uuid: "app-uuid-here", force: true }
Deploy all applications with a tag
coolify_deploy { tag: "production" }
Get application logs
coolify_get_application_logs { uuid: "app-uuid-here", since: 3600 }
Use a specific team
coolify_list_applications { team: "staging" }
Development
# Clone the repository
git clone https://github.com/Softtor/coolify-mcp-server.git
cd coolify-mcp-server
# Install dependencies
npm install
# Build
npm run build
# Run locally
COOLIFY_API_KEY=your-key npm start
Contributing
Contributions are welcome! Please feel free to submit a Pull Request.
License
MIT License - see LICENSE for details.
Related
- Coolify - Self-hostable Heroku/Netlify alternative
- Model Context Protocol - Protocol for AI tool integration
- Claude Code - Anthropic's CLI for Claude
推荐服务器
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。
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。
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。
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。
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。
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。
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。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
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。