coolify-mcp
MCP server for Coolify API that enables full deployment workflows from zero to production, including project, server, and application management.
README
coolify-mcp
MCP server for Coolify API - enables full deployment workflows from zero to production.
Features
- Full Deployment Workflow: Create projects, environments, servers, and applications from scratch
- 6 Application Types: Public git, GitHub App, Deploy Key, Dockerfile, Docker Image, Docker Compose
- Environment Management: Full CRUD for environment variables with secret masking
- Deployment Control: Deploy, start, stop, restart applications
- Security: Write protection, secret redaction, log sanitization
- 38 Tools: Complete coverage of Coolify API operations
Requirements
- Node 18+
- A Coolify API token (Settings > API in your Coolify dashboard)
Install
npm install -g @fndchagas/coolify-mcp
# or
npx -y @fndchagas/coolify-mcp
Quick Start
Claude Code CLI
claude mcp add coolify \
--env COOLIFY_BASE_URL="https://coolify.example.com/api/v1" \
--env COOLIFY_TOKEN="<token>" \
-- npx -y @fndchagas/coolify-mcp
OpenAI Codex CLI
codex mcp add coolify \
--env COOLIFY_BASE_URL="https://coolify.example.com/api/v1" \
--env COOLIFY_TOKEN="<token>" \
-- npx -y @fndchagas/coolify-mcp
Or edit ~/.codex/config.toml:
[mcp_servers.coolify]
command = "npx"
args = ["-y", "@fndchagas/coolify-mcp"]
env = { COOLIFY_BASE_URL = "https://coolify.example.com/api/v1", COOLIFY_TOKEN = "<token>" }
Manual Configuration (~/.mcp.json)
{
"mcpServers": {
"coolify": {
"type": "stdio",
"command": "npx",
"args": ["-y", "@fndchagas/coolify-mcp"],
"env": {
"COOLIFY_BASE_URL": "https://coolify.example.com/api/v1",
"COOLIFY_TOKEN": "<token>",
"COOLIFY_ALLOW_WRITE": "true"
}
}
}
}
Environment Variables
| Variable | Default | Description |
|---|---|---|
COOLIFY_BASE_URL |
required | Coolify API URL (e.g., https://coolify.example.com/api/v1) |
COOLIFY_TOKEN |
required | API token from Coolify Settings > API |
COOLIFY_ALLOW_WRITE |
true |
Enable write operations (create, update, delete, deploy) |
COOLIFY_ALLOW_UNSAFE_LOGS |
false |
Allow raw logs without redaction |
COOLIFY_STRICT_VERSION |
false |
Fail on API version mismatch |
MCP_TRANSPORT |
stdio |
Transport: stdio, http, both |
PORT |
7331 |
HTTP port (when using http transport) |
Deploy from Zero
With this MCP, you can deploy an application from scratch:
1. listProjects / createProject → Get or create a project
2. listEnvironments / createEnvironment → Get or create an environment
3. listServers / createServer → Get or create a server
4. listPrivateKeys / createPrivateKey → Get or create SSH keys (if needed)
5. createPublicApplication → Create the application
6. upsertEnv → Configure environment variables
7. deploy → Trigger deployment
Tools Reference
Projects & Environments
| Tool | Description | Write |
|---|---|---|
listProjects |
List all projects | |
createProject |
Create a new project | ✓ |
updateProject |
Update project name/description | ✓ |
deleteProject |
Delete a project and all its resources | ✓ |
listEnvironments |
List environments in a project | |
createEnvironment |
Create a new environment | ✓ |
Servers & Infrastructure
| Tool | Description | Write |
|---|---|---|
listServers |
List all servers | |
getServer |
Get server details | |
createServer |
Create a new server | ✓ |
validateServer |
Validate server connection | |
listPrivateKeys |
List SSH private keys | |
createPrivateKey |
Create a new SSH key | ✓ |
listGithubApps |
List configured GitHub Apps |
Applications - Read
| Tool | Description |
|---|---|
listApplications |
List all applications (summarized by default) |
getApplication |
Get application details (secrets masked by default) |
getLogs |
Get application runtime logs |
Applications - Create
| Tool | Description | Write |
|---|---|---|
createPublicApplication |
Create from public git repository | ✓ |
createPrivateGithubAppApplication |
Create using GitHub App | ✓ |
createPrivateDeployKeyApplication |
Create using SSH deploy key | ✓ |
createDockerfileApplication |
Create from Dockerfile content | ✓ |
createDockerImageApplication |
Create from Docker image | ✓ |
createDockerComposeApplication |
Create from Docker Compose | ✓ |
Applications - Manage
| Tool | Description | Write |
|---|---|---|
updateApplication |
Update application configuration | ✓ |
deleteApplication |
Delete an application | ✓ |
startApplication |
Start an application | ✓ |
stopApplication |
Stop an application | ✓ |
restartApplication |
Restart an application | ✓ |
Environment Variables
| Tool | Description | Write |
|---|---|---|
listEnvs |
List env vars (secrets masked by default) | |
createEnv |
Create a new env var | ✓ |
upsertEnv |
Create or update env var by key | ✓ |
updateEnv |
Update an existing env var | ✓ |
deleteEnv |
Delete an env var | ✓ |
Deployments
| Tool | Description | Write |
|---|---|---|
deploy |
Trigger a deployment | ✓ |
listDeployments |
List running deployments | |
getDeployment |
Get deployment status and logs | |
listAppDeployments |
List deployments for an application | |
cancelDeployment |
Cancel a running deployment | ✓ |
Databases & Services
| Tool | Description | Write |
|---|---|---|
listDatabases |
List all databases | |
getDatabase |
Get database details | |
listServices |
List one-click services | |
createService |
Create a one-click service | ✓ |
Other
| Tool | Description |
|---|---|
listResources |
List all resources with filtering |
Security Features
Write Protection
Disable all write operations:
COOLIFY_ALLOW_WRITE=false
Secret Masking
- Environment variable values are masked by default
- Database credentials are redacted
- Use
showSecrets: trueonly when necessary
Log Sanitization
Logs are sanitized to remove sensitive data. Control with logMode:
safe(default): Redacts common secret patternsstrict: More aggressive redactionraw: No redaction (requiresCOOLIFY_ALLOW_UNSAFE_LOGS=true)
Development
git clone https://github.com/frndchagas/coolify-mcp.git
cd coolify-mcp
npm install
npm run dev
Scripts
npm run dev # Run in development mode
npm run build # Build TypeScript
npm run generate # Regenerate types from OpenAPI
npm run fetch:openapi # Fetch latest OpenAPI spec
npm run update # Fetch + regenerate
Pinned Coolify Version
Version is defined in src/coolify/constants.ts. To update:
- Edit
COOLIFY_VERSIONinsrc/coolify/constants.ts - Run
npm run update
Registry Listings
- MCP Registry (API): https://registry.modelcontextprotocol.io/v0.1/servers/io.github.frndchagas%2Fcoolify-mcp/versions/0.1.4
- MCP Registry Docs: https://registry.modelcontextprotocol.io/docs
MCP Client Examples
HTTP Client
import { Client } from '@modelcontextprotocol/sdk/client/index.js';
import { StreamableHTTPClientTransport } from '@modelcontextprotocol/sdk/client/streamableHttp.js';
const client = new Client({ name: 'coolify-client', version: '1.0.0' });
const transport = new StreamableHTTPClientTransport(
new URL('http://localhost:7331/mcp')
);
await client.connect(transport);
// List all applications
const apps = await client.callTool({
name: 'coolify.listApplications',
arguments: {},
});
console.log(apps.structuredContent);
// Deploy an application
const deploy = await client.callTool({
name: 'coolify.deploy',
arguments: { uuid: 'your-app-uuid' },
});
console.log(deploy.structuredContent);
await client.close();
Stdio Client
import { Client } from '@modelcontextprotocol/sdk/client/index.js';
import { StdioClientTransport } from '@modelcontextprotocol/sdk/client/stdio.js';
const client = new Client({ name: 'coolify-client', version: '1.0.0' });
const transport = new StdioClientTransport({
command: 'npx',
args: ['-y', '@fndchagas/coolify-mcp'],
env: {
COOLIFY_BASE_URL: 'https://coolify.example.com/api/v1',
COOLIFY_TOKEN: '<token>',
},
});
await client.connect(transport);
const result = await client.callTool({
name: 'coolify.getApplication',
arguments: { uuid: 'your-app-uuid' },
});
console.log(result.structuredContent);
await client.close();
License
MIT
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。