APiGen MCP Server (Java)

Native Java implementation of Model Context Protocol (MCP) server for APiGen - enables AI assistants like Claude to generate production-ready APIs directly from natural language.

Overview

The Problem: Disconnected AI Code Generation

Before APiGen MCP:

User: "Claude, create a REST API for my e-commerce system"

Claude: "Sure! Here's a basic Flask example... [generates incomplete code]
        You'll need to:
        1. Install Flask manually
        2. Set up database manually
        3. Add authentication manually
        4. Configure Docker manually
        5. Write tests manually"

After APiGen MCP:

User: "Generate a REST API for this e-commerce schema:
      CREATE TABLE products (id BIGINT, name VARCHAR(255), price DECIMAL);
      Use java-spring with JWT auth, Docker, and tests"

Claude: [Uses apigen MCP tool]
        ✅ Generated complete Spring Boot project
        ✅ Entities, repositories, services, controllers
        ✅ JWT authentication configured
        ✅ Docker Compose with PostgreSQL
        ✅ Unit + Integration tests
        ✅ OpenAPI documentation
        ✅ Production-ready in 10 seconds

Time Saved: 99% (2-3 days → 5 minutes)

Features

• Direct Integration: Native integration with APiGen codegen (no subprocess calls)
• Full MCP Support: Implements MCP protocol over stdio using JSON-RPC 2.0
• 6 Tools Available: Generate, preview, parse SQL schemas and more
• 4 Resources: Languages, databases, field types, and example schemas

Tools

Tool	Description
generate_from_sql	Generate API code from SQL schema content
preview_sql	Preview generated code without writing files
parse_sql	Parse SQL and return extracted entities
list_languages	List supported programming languages
list_databases	List supported database types
generate_entity	Generate code for a single entity

Resources

URI	Description
apigen://languages	Supported programming languages
apigen://databases	Supported database types
apigen://field-types	Supported field types for entities
apigen://example/sql	Example SQL schema

Configuration

Claude Desktop

Add to your claude_desktop_config.json:

{
  "mcpServers": {
    "apigen": {
      "command": "java",
      "args": ["-jar", "/path/to/apigen/mcp/java/build/libs/java-1.0.0-SNAPSHOT-all.jar"]
    }
  }
}

Comparison with Python MCP

Feature	Java MCP	Python MCP
Integration	Direct (native)	Subprocess (CLI)
Startup time	Slower (JVM)	Faster
Performance	Better for heavy generation	Good
Dependencies	Embedded	Requires Java CLI

Choose Java MCP for:

• Production deployments
• Heavy code generation workloads
• When you want native integration

Choose Python MCP for:

• Quick setup
• Lightweight usage
• When you already have APiGen CLI installed