Skip to main content
Glama
vltansky

MCP Server Boilerplate

by vltansky
OverviewSchemaSourceRelated ServersDiscussionsAdmin
TypeScript

MCP Server Boilerplate

A TypeScript template for building Model Context Protocol (MCP) servers.

This boilerplate provides a solid foundation for creating MCP servers that can integrate with Cursor, Claude, and other AI assistants. It includes best practices, example tools, proper error handling, and a well-structured TypeScript codebase.

What This Template Provides

  • Complete MCP Server Setup: Ready-to-use server with proper configuration

  • Example Tools: Demonstrates common MCP tool patterns and best practices

  • TypeScript Integration: Full type safety with Zod validation

  • Error Handling: Robust error handling patterns throughout

  • Testing Setup: Vitest configuration for unit testing

  • Development Workflow: Build, watch, and inspection scripts

Related MCP server: JSON MCP Boilerplate

Key Features

Type-Safe Development: Built with TypeScript and Zod for runtime validation and compile-time safety.

Modular Architecture: Well-organized code structure with separate modules for tools, utilities, and types.

Example Patterns: Demonstrates data retrieval, search, analytics, and system utilities.

Development Ready: Includes hot reload, testing, and MCP inspector integration.

Quick Start

1. Clone and Setup

git clone https://github.com/vltansky/mcp-boilerplate.git
cd mcp-server-boilerplate
yarn install
yarn build

2. Configure MCP Client

Add to your .cursor/mcp.json or other MCP client configuration:

{
  "mcpServers": {
    "my-custom-server": {
      "command": "node",
      "args": ["path/to/your/dist/server.js"]
    }
  }
}

3. Start Developing

yarn watch  # Start development with hot reload

4. Test Your Tools

Use the MCP inspector to test your tools:

yarn inspector

Task Master - Getting Started

Once you have your MCP server running, here's how to get the most out of the Task Master workflow:

Next Steps for Project Success

  1. Configure AI models (if needed) and add API keys to .env

    • Models: Use task-master models commands

    • Keys: Add provider API keys to .env (or inside the MCP config file i.e. .cursor/mcp.json)

  2. Discuss your idea with AI and ask for a PRD using example_prd.txt, and save it to scripts/PRD.txt

  3. Ask Cursor Agent (or run CLI) to parse your PRD and generate initial tasks:

    • MCP Tool: parse_prd | CLI: task-master parse-prd scripts/prd.txt

  4. Ask Cursor to analyze the complexity of the tasks in your PRD using research

    • MCP Tool: analyze_project_complexity | CLI: task-master analyze-complexity

  5. Ask Cursor to expand all of your tasks using the complexity analysis

  6. Ask Cursor to begin working on the next task

  7. Add new tasks anytime using the add-task command or MCP tool

  8. Ask Cursor to set the status of one or many tasks/subtasks at a time. Use the task id from the task lists.

  9. Ask Cursor to update all tasks from a specific task id based on new learnings or pivots in your project.

  10. Ship it!

Example Tools Included

Core Tools

  • get_data - Demonstrates data retrieval with filtering and pagination

  • search_items - Shows search functionality with multiple search types (exact, fuzzy, regex)

  • analyze_data - Example analytics tool with chart data generation

  • get_system_info - System utilities for date, timezone, and version information

Tool Patterns Demonstrated

  • Parameter Validation: Using Zod schemas for type-safe input validation

  • Error Handling: Consistent error handling and user-friendly error messages

  • Async Operations: Proper async/await patterns with timeout simulation

  • Response Formatting: JSON and compact-JSON output modes

  • Type Safety: Full TypeScript integration with proper type inference

Project Structure

src/
├── server.ts              # Main MCP server setup and tool registration
├── tools/
│   └── example-tools.ts    # Example tool implementations
└── utils/
    └── formatter.ts        # Response formatting utilities

docs/                       # Documentation files
package.json               # Dependencies and scripts
tsconfig.json              # TypeScript configuration
vitest.config.ts           # Testing configuration

Customizing for Your Use Case

1. Replace Example Tools

Edit src/tools/example-tools.ts to implement your business logic:

export async function yourCustomOperation(input: YourInputType): Promise<YourOutputType> {
  // Your implementation here
  return result;
}

2. Update Server Registration

Modify src/server.ts to register your tools:

server.tool(
  'your_tool_name',
  'Description of what your tool does',
  {
    // Zod schema for parameters
    param1: z.string().describe('Parameter description'),
    param2: z.number().optional().default(10)
  },
  async (input) => {
    // Tool implementation
    const result = await yourCustomOperation(input);
    return {
      content: [{
        type: 'text',
        text: formatResponse(result, input.outputMode)
      }]
    };
  }
);

3. Add Your Data Layer

Create modules for your specific data sources:

src/
├── database/          # Database connections and queries
├── external-apis/     # External API integrations
├── file-system/       # File system operations
└── your-domain/       # Your business logic

Development Workflow

Available Scripts

  • yarn build - Compile TypeScript to JavaScript

  • yarn watch - Watch mode for development

  • yarn start - Run the compiled server

  • yarn test - Run unit tests

  • yarn test:ui - Run tests with UI

  • yarn inspector - Start MCP inspector for testing tools

Testing Your Tools

  1. Unit Tests: Add tests alongside your tool files

  2. Integration Testing: Use the MCP inspector to test tool behavior

  3. Manual Testing: Test with actual MCP clients like Cursor

Adding Dependencies

For data sources, add appropriate dependencies:

# Database
yarn add sqlite3 @types/sqlite3

# HTTP requests
yarn add axios

# File processing
yarn add fs-extra @types/fs-extra

# Date handling
yarn add date-fns

MCP Best Practices

Tool Design

  • Clear Descriptions: Write detailed tool descriptions for AI assistants

  • Parameter Validation: Use Zod for runtime validation

  • Error Handling: Provide meaningful error messages

  • Output Consistency: Use consistent response formats

Performance

  • Async Operations: Use async/await for all I/O operations

  • Resource Management: Clean up resources properly

  • Caching: Implement caching for expensive operations

  • Pagination: Support pagination for large datasets

Security

  • Input Validation: Validate all inputs with Zod

  • Error Messages: Don't expose sensitive information in errors

  • Resource Limits: Implement appropriate limits and timeouts

  • Authentication: Add authentication if accessing sensitive data

Common Use Cases

File System Tools

  • File search and indexing

  • Content analysis

  • Code parsing and analysis

Database Integration

  • Query interfaces

  • Data analysis and reporting

  • Schema exploration

External API Integration

  • API wrapping and simplification

  • Data aggregation from multiple sources

  • Rate limiting and caching

Development Tools

  • Code generation

  • Testing utilities

  • Build and deployment helpers

Contributing

  1. Fork this repository

  2. Create your feature branch

  3. Add tests for new functionality

  4. Ensure all tests pass

  5. Submit a pull request

License

MIT License - feel free to use this template for your own projects.

Resources

Ready to build your MCP server? Start by customizing the example tools in src/tools/example-tools.ts and updating the server registration in src/server.ts.

Available Tools

2 tools
get_dataB

Retrieve data from your custom data source with optional filtering and pagination.

ParametersJSON Schema
NameRequiredDescriptionDefault
limitNoMaximum number of items to return (1-100)
filterNoFilter criteria for the data
outputModeNoOutput format: "json" for formatted JSON (default), "compact-json" for minified JSONjson
includeMetadataNoInclude additional metadata in the response

TDQS

B3.1/5.0
Behavior2/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

No annotations are provided, so the description carries the full behavioral disclosure burden. It does not mention authentication requirements, rate limits, or what 'metadata' in the schema means. The description adds no behavioral context beyond what the schema already states.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness4/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The description is a single sentence, front-loaded with the action ('Retrieve data'), and wastes no words. However, 'custom data source' is under-specified, which slightly reduces clarity.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness2/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

With no output schema and no annotations, the description should elaborate on return values and behavioral context. It fails to mention the meaning of 'metadata', the nature of the data source, or potential error conditions, making it incomplete for a 4-parameter tool.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters3/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

The input schema describes all four parameters with clear descriptions, so baseline is 3. The description's mention of 'filtering and pagination' adds minimal extra meaning by hinting at the filter and limit parameters, but does not explain outputMode or includeMetadata beyond the schema.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose4/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description uses the specific verb 'Retrieve' with the object 'data from your custom data source', clearly indicating the tool's function. It implicitly distinguishes from the sibling 'get_system_info' by focusing on custom data, but 'custom data source' is vague and lacks detail.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines3/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

The phrase 'with optional filtering and pagination' implies when to use the tool (when you need data with these capabilities), but there is no explicit guidance on alternatives or when not to use it. No comparison with the sibling tool 'get_system_info' is provided.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

get_system_infoA

Get system information and utilities. Provides current date, timezone, and other helpful context.

ParametersJSON Schema
NameRequiredDescriptionDefault
infoNoType of system information to retrieveall

TDQS

A3.8/5.0
Behavior3/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

With no annotations provided, the description must carry the burden of behavioral disclosure. It states the tool provides information, implying a read-only action, but does not mention output format, version specifics, or any potential side effects. This is adequate but not rich.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness5/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The description is concise, using two short sentences to convey the core purpose. It is front-loaded with the key verb and resource, with no superfluous information.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness4/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a simple tool with one optional parameter and no output schema, the description is largely complete. It covers the main use cases (date, timezone) and implies additional context, though it does not explicitly mention the 'info' parameter or the 'version' option. The rich schema compensates for this.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters3/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

The input schema has 100% coverage with a clear description for the 'info' parameter and an enum defining valid values. The description adds no additional parameter semantics beyond what the schema already provides, so the baseline of 3 is appropriate.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description clearly states the tool's purpose: getting system information such as date and timezone. It uses a specific verb and resource, and is distinct from the sibling tool 'get_data'.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines3/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

The description implies usage for retrieving system context, but it does not explicitly differentiate from get_data or explain when to use this tool over alternatives. No exclusion criteria are provided.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

  1. 2 tool updatesv0.1.0
    • First observedget_data
    • First observedget_system_info
Deploy Server
RayrunBuild, deploy, and operate MCP servers without becoming an infrastructure team.ray.runAd

TDQS

B3.4/5.0

Scored across 2 tools

Disambiguation5/5

The two tools serve entirely different purposes: one retrieves data, the other provides system information. There is no overlap or ambiguity, so an agent can easily distinguish them.

Naming Consistency5/5

Both tool names follow the same 'get_' prefix followed by a noun, resulting in a consistent naming convention.

Tool Count3/5

With only two tools, the set feels minimal and thin. While it is a boilerplate, the number is at the lower edge of the typical range, making it borderline.

Completeness2/5

The tool surface is very limited: it offers only retrieval operations for data and a single system info utility. There are no create/update/delete operations for data, and the tools do not form a coherent workflow, leaving significant gaps for real use.

Maintenance

ActivityInactive
ResponsivenessNo issues

Resources

Related MCP Connectors

Related MCP Servers

  • JSON MCP Boilerplate

    ricleedo
    F
    license
    A
    quality
    D
    maintenance
    A starter template for building MCP (Model Context Protocol) servers with TypeScript support. Provides a clean foundation with example tools, resources, and prompts for creating custom integrations with Claude, Cursor, or other MCP-compatible AI assistants.
    2
    6
    -

  • MCP Server Boilerplate

    ricleedo
    F
    license
    C
    quality
    D
    maintenance
    A starter template for building MCP (Model Context Protocol) servers that integrate with Claude, Cursor, or other MCP-compatible AI assistants. Provides a clean foundation with TypeScript support, example tool implementation, and installation scripts for quick customization.
    2
    6
    -

  • MCP Server Starter

    TheSethRose
    A
    license
    Not graded
    quality
    D
    maintenance
    A TypeScript template for building Model Context Protocol (MCP) servers that enables developers to quickly create custom tools and integrate them with AI platforms like Claude, Cursor, Windsurf, and Cline.
    7
    34
    MIT