MCP Todo Server
A demo Todo application server built with a clean architecture using MCPServer and JSON Placeholder.
mcp-todo-server/README.md
MCP Todo Server
This project is a demonstration of a clean architecture implementation using Node.js and TypeScript. It utilizes the MCPServer from the @modelcontextprotocol/sdk to create a server and manage todos through a JSON placeholder API.
Demo in MCP Inspector

Demo in Cursor

Technology Stack
This project is built with modern technologies and follows clean architecture principles:
Core Technologies
- Node.js: Runtime environment for executing JavaScript code server-side
- TypeScript: Strongly typed programming language that builds on JavaScript
- Model Context Protocol (MCP): Protocol for AI models to interact with external tools and data sources
Backend Framework
- @modelcontextprotocol/sdk: Official SDK for creating MCP-compliant servers
- Zod: TypeScript-first schema validation library used for input validation
Testing
- Jest: JavaScript testing framework
- Supertest: HTTP assertions library for testing API endpoints
Development Tools
- tsx: TypeScript execution environment with native ESM support
- tsc-alias: Tool for resolving TypeScript path aliases during compilation
- ts-node: TypeScript execution engine for Node.js
Architecture
- Clean Architecture: The project follows a clean architecture approach with:
- Domain layer: Core business logic and entities
- Application layer: Use cases and application services
- Infrastructure layer: External interactions and implementations
- Presentation layer: API endpoints and request/response handling
External APIs
- JSONPlaceholder: RESTful API for testing and prototyping providing fake todo data
Architecture Diagram
Sequence Diagram
The following diagram shows how requests flow through the MCP Todo Server:
sequenceDiagram
participant Client as MCP Client/Inspector
participant Server as McpServer
participant Handler as TodoHandlers
participant Service as TodoService
participant Repository as JsonPlaceholderTodoRepository
participant API as JSONPlaceholder API
Client->>Server: Tool Call Request (e.g., get_todos)
alt Get All Todos
Server->>Service: todoService.getAllTodos()
Service->>Repository: todoRepository.findAll()
Repository->>API: fetch('https://jsonplaceholder.typicode.com/todos')
API-->>Repository: JSON Response
Repository-->>Service: Todo[] entities
Service-->>Server: Todo[] entities
Server-->>Client: Formatted JSON Response
else Get Todo by ID
Client->>Server: get_todo_by_id with id parameter
Server->>Service: todoService.getTodoById(id)
Service->>Repository: todoRepository.findById(id)
Repository->>API: fetch('https://jsonplaceholder.typicode.com/todos/{id}')
API-->>Repository: JSON Response
Repository-->>Service: Todo entity
Service-->>Server: Todo entity
Server-->>Client: Formatted JSON Response
else Create Todo
Client->>Server: create_todo with title, completed
Server->>Service: todoService.createTodo(data)
Service->>Repository: todoRepository.create(data)
Repository->>API: POST fetch('https://jsonplaceholder.typicode.com/todos')
API-->>Repository: JSON Response
Repository-->>Service: New Todo entity
Service-->>Server: New Todo entity
Server-->>Client: Formatted JSON Response
else Update Todo
Client->>Server: update_todo with id, title, completed
Server->>Service: todoService.updateTodo(id, data)
Service->>Repository: todoRepository.update(id, data)
Repository->>API: PUT fetch('https://jsonplaceholder.typicode.com/todos/{id}')
API-->>Repository: JSON Response
Repository-->>Service: Updated Todo entity
Service-->>Server: Updated Todo entity
Server-->>Client: Formatted JSON Response
else Delete Todo
Client->>Server: delete_todo with id
Server->>Service: todoService.deleteTodo(id)
Service->>Repository: todoRepository.delete(id)
Repository->>API: DELETE fetch('https://jsonplaceholder.typicode.com/todos/{id}')
API-->>Repository: Response status
Repository-->>Service: Boolean result
Service-->>Server: Boolean result
Server-->>Client: Success/Failure message
end
Project Structure
mcp-todo-server
├── src
│ ├── domain
│ │ ├── entities
│ │ │ └── todo.ts
│ │ ├── repositories
│ │ │ └── todoRepository.ts
│ │ └── valueObjects
│ │ └── todoId.ts
│ ├── application
│ │ ├── services
│ │ │ └── todoService.ts
│ │ └── useCases
│ │ ├── createTodo.ts
│ │ ├── deleteTodo.ts
│ │ ├── getTodoById.ts
│ │ ├── getTodos.ts
│ │ └── updateTodo.ts
│ ├── infrastructure
│ │ ├── repositories
│ │ │ └── jsonPlaceholderTodoRepository.ts
│ │ └── http
│ │ └── httpClient.ts
│ ├── presentation
│ │ └── handlers
│ │ └── todoHandlers.ts
│ ├── server.ts
│ └── index.ts
├── tests
│ ├── unit
│ │ ├── domain
│ │ │ └── entities
│ │ │ └── todo.test.ts
│ │ ├── application
│ │ │ └── services
│ │ │ └── todoService.test.ts
│ │ └── infrastructure
│ │ └── repositories
│ │ └── jsonPlaceholderTodoRepository.test.ts
│ └── integration
│ └── server.test.ts
├── http
│ └── todo-api.http
├── package.json
├── tsconfig.json
├── jest.config.js
└── README.md
Setup Instructions
-
Clone the repository:
git clone <repository-url> cd mcp-todo-server -
Install dependencies:
npm install -
Run the server:
npm start
Usage
The server exposes various endpoints for managing todos. You can use the provided http/todo-api.http file to test the API endpoints manually.
Using MCP Inspector
What is MCP Inspector?
MCP Inspector is a tool that allows you to interact with MCP-compliant servers directly, testing tool calls and viewing responses without needing to integrate with an AI model.
Installation
To install MCP Inspector globally:
npm install -g @modelcontextprotocol/inspector
Connecting to MCP todo Server
# 1. First, build and start MCP Todo server:
npm run build
npm start
# 2. In a separate terminal, run MCP Inspector and connect it to mcp todo server:
mcp-inspector --server "node ./build/index.js"
or
npx @modelcontextprotocol/inspector node build/index.js
# Alternatively, if mcp todo server is already running, can pipe it to the inspector:
node ./build/index.js | mcp-inspector
# 3. Once connected, the inspector will open in your default web browser, allowing you to:
# Browse available tools
# Execute tool calls with custom parameters
# View responses in a formatted JSON view
# Debug request/response cycles
Setup MCP Todo Server in Cursor
-
Configure MCP integration:
-
Open Cursor settings
-
Navigate to the Extensions or AI Tools section
-
Find "Model Context Protocol" or "MCP Tools" settings mcp.json
{
"mcpServers": {
... other mcp server,
"todo-mcp-server": {
"command": "node",
"args": [
"D:\\dev\\node\\mcp-todo-server\\build\\index.js"
]
}
}
}
Testing
To run the unit and integration tests, use the following command:
npm test
License
This project is licensed under the MIT License.
Máy chủ liên quan
Scout Monitoring MCP
nhà tài trợPut performance and error data directly in the hands of your AI assistant.
Alpha Vantage MCP Server
nhà tài trợAccess financial market data: realtime & historical stock, ETF, options, forex, crypto, commodities, fundamentals, technical indicators, & more
Model Context Protocol servers
A collection of reference implementations for the Model Context Protocol (MCP), demonstrating secure and controlled access to tools and data sources for Large Language Models (LLMs).
Lerian MCP Server
Provides educational content, model information, and read-only API interactions for Lerian developers.
Nuanced MCP Server
Provides call graph analysis for LLMs using the nuanced library.
MCP Server Pentest
Automated browser penetration testing to detect XSS and SQL vulnerabilities.
Narsil MCP
Blazingly fast 🔥 best in class MCP server in Rust 🦀 with neural engine, security profiling, and optional graph frontend
mistaike.ai
MCP security gateway with DLP scanning (PII, secrets, API keys), prompt injection protection, Memory Vault, Bug Vault (295k+ patterns), and unified audit logging. Two endpoints: free bug search at /mcp and authenticated hub at /hub_mcp.
AST2LLM for Go
An AST-powered tool that enhances LLM context by automatically injecting relevant Go code structures into prompts.
vigile-mcp
Security scanner for MCP servers and agent skills — query trust scores, check for vulnerabilities, and search the Vigile trust registry
VoteShip
MCP server for VoteShip - manage feature requests, votes, roadmaps, and changelogs from any MCP client. 22 tools, 5 resources, 4 workflow prompts. Triage feedback, detect duplicates, plan sprints, and generate changelogs with AI.
Raymon
Stateful HTTP ingest + MCP server + terminal UI for Ray-style logs.