Skip to content
MCP server

MetaEngine

Generate typed code in 11 languages from JSON recipes, type graphs and API schemas

Runs locallyStatus and history
Pricing
Open source
Platforms
CLI, API, Linux
Runs
Locally
Version
1.5.0
Listing updated
today

About MetaEngine

MetaEngine is an open-source MCP server for developer tools. It runs locally on your machine and is started by your AI client.

MetaEngine takes a description of types rather than a prompt for code. You supply a compact JSON recipe, a full type graph, or an OpenAPI, GraphQL, Protocol Buffers or SQL DDL source, and the server writes related source files to a local folder. Explicit references between types let the generator resolve names and imports across a whole batch, and hand-written code can be attached through customCode.

Recipes keep repeated structure short: a trailing question mark marks an optional field, a ref points at another type, and named templates with parameter rows expand locally before anything is sent. Unknown fields, unresolved references and duplicate names fail on your machine before the hosted request. Generation itself runs on the MetaEngine API, which the site says processes payloads ephemerally without storing or logging them, and needs no key or account.

Each request is limited to 250 counted types and 5 MiB of JSON. Existing files are skipped by default, and a dry run returns the output without writing it. Output covers TypeScript, Python, Go, C#, Java, Kotlin, Groovy, Scala, Swift, PHP and Rust, with typed clients for frameworks such as Angular, React and Spring.

Install

Run in a terminal

claude mcp add mcp-server -- npx -y @metaengine/mcp-server

Add to claude_desktop_config.json

{
  "mcpServers": {
    "mcp-server": {
      "command": "npx",
      "args": [
        "-y",
        "@metaengine/mcp-server"
      ]
    }
  }
}
Install in CursorOpens Cursor, which asks you to confirm. Or add it by hand:

Add to ~/.cursor/mcp.json

{
  "mcpServers": {
    "mcp-server": {
      "command": "npx",
      "args": [
        "-y",
        "@metaengine/mcp-server"
      ]
    }
  }
}
Install in VS CodeOpens VS Code, which asks you to confirm. Or add it by hand:

Add to .vscode/mcp.json

{
  "servers": {
    "mcp-server": {
      "type": "stdio",
      "command": "npx",
      "args": [
        "-y",
        "@metaengine/mcp-server"
      ]
    }
  }
}

Add to ~/.codeium/windsurf/mcp_config.json

{
  "mcpServers": {
    "mcp-server": {
      "command": "npx",
      "args": [
        "-y",
        "@metaengine/mcp-server"
      ]
    }
  }
}

Add to your client's MCP settings (most use this shape)

{
  "mcpServers": {
    "mcp-server": {
      "command": "npx",
      "args": [
        "-y",
        "@metaengine/mcp-server"
      ]
    }
  }
}

Configuration

  • METAENGINE_API_URLOverride the MetaEngine generation API endpoint (defaults to https://api.metaengine.eu)

Tools this server provides (8)

Taken from the project documentation; the running server may differ.

generate_from_recipe
Expand compact JSON recipes into a type graph and generate files.
generate_code
Generate from a complete native JSON type graph.
load_spec_from_file
Read a native generate_code payload from a local JSON file.
generate_openapi
Generate a typed client from an OpenAPI document.
generate_graphql
Generate a typed client from GraphQL SDL.
generate_protobuf
Generate a typed client from Protocol Buffers definitions.
generate_sql
Generate model types from SQL DDL.
metaengine_initialize
Load generation rules and language notes for the assistant.

What it does

Recipe generation
Expand compact JSON recipes with templates into full type graphs, then generate files.
Native type graphs
Generate from a complete graph including custom target-language code.
API client generation
Typed clients from OpenAPI, GraphQL SDL and Protocol Buffers.
SQL model types
Create model types from SQL DDL.
Safe file writing
Skips existing files by default and supports dry runs.
Local validation
Resolves references and limits before the hosted call.

Good for

  • Regenerate an API client after a schema change
  • Produce models for several languages from one definition
  • Turn SQL DDL into typed models
  • Keep version-controlled recipe files that rebuild models on demand

Pricing

Open source

MIT-licensed server; hosted generation needs no account, paid plans not mentioned

Prices change. Check metaengine.eu before you buy.

MetaEngine FAQ

Do I need an API key?

No, the README says the hosted service requires no key or account.

What is the size limit?

250 counted types per request and a 5 MiB JSON limit for the recipe plus input files.

Is my code uploaded?

Only the generation payload is sent; existing project source is not uploaded unless included in it.

Which Node version?

Node.js 18 or later.

Is MetaEngine free?

Yes. MetaEngine is open source, so you can use and self-host it for free.

How do I add MetaEngine to Claude Code?

Run this command in your terminal: claude mcp add mcp-server -- npx -y @metaengine/mcp-server

What are the best alternatives to MetaEngine?

The closest alternatives to MetaEngine are REA, Codebase Memory, GitHub, Serena MCP: the IDE for your agent and XERJ.

Alternatives to MetaEngine

  1. REA

    MCP server

    CLI and MCP server that lets coding agents reverse engineer apps and binaries

    Runs locally
    • Open source
    • 464K GitHub stars
    • Developer tools
  2. Codebase Memory

    MCP server

    Local MCP server that builds a code knowledge graph for AI coding agents

    Runs locally
    • Open source
    • 48.7K GitHub stars
    • Knowledge and memory
  3. GitHub

    MCP server

    Official MCP server connecting AI assistants to GitHub repos, issues and workflows

    Not checked yet
    • Remote
    • Open source
    • 34.5K GitHub stars
    • Developer tools
  4. MCP toolkit giving coding agents symbol-level code retrieval and editing

    Runs locally
    • Open source
    • 31.5K GitHub stars
    • Developer tools
  5. XERJ

    MCP server

    Local search engine for AI agents with cited answers over your code and docs

    Runs locally
    • Open source
    • 19.2K GitHub stars
    • Knowledge and memory
  6. Ouroboros

    MCP server

    Spec-first layer that pins requirements and verifies results of AI coding agent work

    Runs locally
    • Open source
    • 6.4K GitHub stars
    • Developer tools