MetaEngine

npm version License: MIT MCP

Code generation, handed to your agent.

MetaEngine exposes its code-generation platform — spec converters for OpenAPI, GraphQL, Protobuf, and SQL, plus a batch type generator — as Model Context Protocol tools. Connect the server to Claude Code, Claude Desktop, Cursor, Cline, or any MCP-aware assistant, and "regenerate my billing client from the new OpenAPI spec" becomes a real, typed, ready-to-commit diff.

Listed on the official MCP Registry as eu.metaengine/mcp-server.

---

Quick Links

• npm package: @metaengine/mcp-server
• Website & docs: metaengine.eu/mcp
• Playground: metaengine.eu/playground

---

Installation

Claude Code:

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

Claude Desktop, Cursor, Cline, or any other MCP client — add to the client's MCP config (claude_desktop_config.json, .cursor/mcp.json, …):

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

That's it. No API key, no signup, free to use.

---

Tools

Seven tools. One call each. Plain text back.

Tool	What it does
generate_openapi	Typed HTTP client from an OpenAPI 3.x document, passed inline or by URL — 10 frameworks
generate_graphql	Typed client from a GraphQL SDL schema, optionally with reusable named fragments — 10 frameworks
generate_protobuf	Typed client from Protocol Buffers (.proto) definitions — 10 frameworks
generate_sql	Typed model classes from SQL DDL (CREATE TABLE), parsed dialect-agnostically — 11 languages
generate_code	Arbitrary type graphs (classes, interfaces, enums, generics) from one structured spec, with imports and cross-references resolved — 11 languages
load_spec_from_file	Runs a generate_code spec from disk, so multi-file architectures stay version-controlled and context usage drops to a file path
metaengine_initialize	Primes the agent before its first generation: patterns, examples, and language-specific rules

Every call is stateless and self-contained: pass the spec inline (or by file path), pick a framework or language, get a write summary back as text. dryRun returns the generated contents inline instead of writing — ready to diff. skipExisting (default) protects files you've already customized.

---

Spec-first development

Your specs are already the source of truth — the OpenAPI document, the GraphQL schema, the .proto files, the DDL. This server puts them to work inside the agent loop: when a spec changes, the agent regenerates the typed surface instead of hand-editing it.

• 4 source specs — OpenAPI 3.x, GraphQL SDL, Protocol Buffers, SQL DDL
• 10 client frameworks — Angular, React, TypeScript Fetch, Go net/http, Java Spring, Python httpx, C# HttpClient, Kotlin Ktor, Rust Reqwest, Swift URLSession
• 11 languages for type and model generation — TypeScript, Python, Go, C#, Java, Kotlin, Groovy, Scala, Swift, PHP, Rust — each emitted idiomatically (data classes in Kotlin, case classes in Scala, structs in Swift and Rust)
• Deterministic — generation is byte-reproducible at a fixed engine version, so agents can retry without drift

The converters surfaced through MCP are the same compiler pipeline that powers the MetaEngine Playground: a spec is parsed, normalized to MetaEngine's intermediate representation, and emitted through a language-specific target. Versions stay in lockstep across surfaces.

For small tasks — a handful of files, exploratory code, one-off scripts — an agent's direct generation is simpler, and agents are told exactly that. The server earns its place when the work is spec-driven, polyglot, or structurally repetitive.

---

Measured behavior in agent loops

Agents that batch through this MCP run with substantially fewer turns and lower cumulative context re-reads than a file-by-file Write loop (~5 turns vs ~75 for the same DDD codebase).

For reproducible measurements across languages, models, and spec shapes, see benchmark/ — a self-contained harness with the prompts, judging tools, and 15 canonical result folders. Numbers there are illustrations from one author's runs at N=5 per cell; reproduce in your own environment to see what holds for you.

---

Context Durability

In long-running sessions where context may be summarized (compaction), MetaEngine survives in three ways:

• Short loop by design — the MCP returns many files per call rather than per turn, so the conversation stays small enough that compaction is rarely triggered (~5 turns vs ~75 for file-by-file Write — see benchmark for measurements).
• Recovery path — the full AI guide is embedded in the tool description on first use; after a successful call, the description swaps to a short directive that points the assistant back at metaengine_initialize, which returns the guide content directly. If compaction wipes the guide, the breadcrumb is enough to reload it.
• Disk-backed state — when the spec is loaded via load_spec_from_file, it lives outside the conversation. A compacted (or fully reset) session can re-run the producing script and pick up without re-reading anything.

---

Documentation

The AI guide is automatically embedded in the tool description on first use — no manual reading required. For reference:

• METAENGINE_AI_GUIDE.md — Critical rules, patterns, language notes, and common mistakes
• EXAMPLES.md — Real-world usage with input/output across all languages

---

Privacy & Pricing

• Free — no API key, no signup, unlimited requests
• Private — specs sent for generation are never saved or logged (see PRIVACY.md)
• Local — MCP server runs on your machine over stdio, MIT licensed
• Terms — See TERMS.md for usage terms

---

Support

• Issues: GitHub Issues
• Email: [email protected]
• Website: metaengine.eu

---

License

MIT License - see LICENSE file for details.

---

About This Repository

This is the documentation and issue tracking repository for MetaEngine MCP Server. The compiled NPM package is available at @metaengine/mcp-server.

Source code is proprietary, but the MCP server is free to use under MIT license.