AutotaskMCP
Enables interaction with Autotask REST API for ticket management and querying through natural language.
README
AutotaskMCP Server
MCP Server built with the MCP Builder [https://mcpbuilder.leniolabs.com] using the Streamable HTTP transport over ExpressJS.
Description
This MCP (Model Context Protocol) server provides tools for interacting with Autotask REST API services. It uses Express.js with Streamable HTTP transport to handle MCP requests.
Features
- Streamable HTTP transport for MCP protocol
- Express.js server for handling HTTP requests
- Multiple Autotask API tools for ticket management and querying
- Session management with proper cleanup
Prerequisites
- Node.js 20 or higher
- npm or yarn
Installation
npm install
Development
npm run dev
This will start the server in watch mode using tsx.
Building
npm run build
This compiles TypeScript to JavaScript in the dist/ directory.
Running
npm start
The server will start on port 3000 by default.
Docker
Build the Docker image
docker build -t autotaskmcp:latest .
Run the container
docker run -p 3000:3000 autotaskmcp:latest
Using Docker Compose
docker-compose up -d
Environment Variables
PORT- Port number for the server (default: 3000)NODE_ENV- Node environment (default: production in Docker)
API Endpoints
POST /mcp- Handle MCP requests (initialization and subsequent requests)GET /mcp- SSE stream for MCP responsesDELETE /mcp- Terminate MCP session
Configuration
Autotask API Credentials
The server requires Autotask API credentials to be configured. These can be set via environment variables:
AUTOTASK_API_INTEGRATION_CODE- Your Autotask API Integration CodeAUTOTASK_USER_NAME- Your Autotask API UsernameAUTOTASK_SECRET- Your Autotask API SecretAUTOTASK_IMPERSONATION_RESOURCE_ID- Your Autotask Impersonation Resource ID
These can be set:
- As environment variables at runtime (recommended)
- As build arguments during Docker build (baked into image)
Deployment
Deploying to Northflank
-
Build Context: Set to
.(the root directory where the Dockerfile is located) -
Port: Expose port
3000to the internet -
Environment Variables: Set the following environment variables in Northflank:
AUTOTASK_API_INTEGRATION_CODE- Your Autotask API Integration CodeAUTOTASK_USER_NAME- Your Autotask API UsernameAUTOTASK_SECRET- Your Autotask API SecretAUTOTASK_IMPERSONATION_RESOURCE_ID- Your Autotask Impersonation Resource IDPORT(optional) - Server port, defaults to 3000NODE_ENV(optional) - Set toproduction
-
Build Arguments (Optional - if you want to hardcode credentials into the image):
AUTOTASK_API_INTEGRATION_CODEAUTOTASK_USER_NAMEAUTOTASK_SECRETAUTOTASK_IMPERSONATION_RESOURCE_ID
Note: If using build arguments, the credentials will be baked into the Docker image. If using environment variables, they can be changed without rebuilding.
Publishing
Publishing to GitHub
-
Initialize Git repository (if not already done):
git init git add . git commit -m "Initial commit" -
Create a GitHub repository and push:
git remote add origin https://github.com/YOUR_USERNAME/autotaskmcp.git git branch -M main git push -u origin main
Publishing Docker Images (Optional)
Note: If you're deploying to Northflank or similar platforms that build from source, you don't need to publish images to a registry. The platform will build directly from your Dockerfile.
Option 1: GitHub Container Registry (Default - Automated)
The repository includes GitHub Actions workflows that automatically build and push to GitHub Container Registry (ghcr.io):
-
Push to main/master branch:
- The CI workflow automatically builds and pushes on every push
- Images are published to:
ghcr.io/YOUR_USERNAME/autotaskmcp:latest - No additional setup required (uses GitHub token automatically)
-
Create a release:
- Create a GitHub release to trigger versioned tags
- The
docker-publish.ymlworkflow handles release builds
Option 2: Manual Build and Push to GitHub Container Registry
-
Build the Docker image:
docker build -t ghcr.io/YOUR_USERNAME/autotaskmcp:latest . -
Login to GitHub Container Registry:
echo $GITHUB_TOKEN | docker login ghcr.io -u YOUR_USERNAME --password-stdin -
Push the image:
docker push ghcr.io/YOUR_USERNAME/autotaskmcp:latest
Option 3: Docker Hub (Optional - Manual Setup Required)
If you prefer Docker Hub, you can manually build and push:
-
Build the Docker image:
docker build -t YOUR_DOCKERHUB_USERNAME/autotaskmcp:latest . -
Login to Docker Hub:
docker login -
Push the image:
docker push YOUR_DOCKERHUB_USERNAME/autotaskmcp:latest
Note: The GitHub Actions workflows use GitHub Container Registry by default. To use Docker Hub with GitHub Actions, you would need to:
- Add
DOCKER_USERNAMEandDOCKER_PASSWORDsecrets to your repository - Modify the workflows to use Docker Hub instead of GHCR
License
MIT
Author
MCP Builder https://mcpbuilder.leniolabs.com
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。