Meta MCP Server

The original MCP server that creates other MCP servers. Built before Anthropic shipped their own mcp-builder skill -- now modernized to align with MCP spec 2025-11-25 and SDK v1.27+.

What It Does

meta-mcp-server exposes tools that let an LLM (or any MCP client) scaffold, write, validate, and template new MCP server projects. Instead of manually setting up package.json, tsconfig.json, and boilerplate tool registrations, you describe what you want and meta-mcp-server writes it.

Related MCP server: MCP-YNU FastMCP Server

Tools

All tools use explicit JSON Schema (via Zod), tool annotations (readOnlyHint, destructiveHint, idempotentHint, openWorldHint), and structured error responses for model self-correction.

Requirements

• Node.js >= 18

• npm

Installation

git clone https://github.com/DMontgomery40/meta-mcp-server.git
cd meta-mcp-server
npm install
npm run build

Usage

stdio transport (default)

npm start
# or
node build/main.js

Streamable HTTP transport

npm run start:http
# or
node build/main.js --http
# Listens on http://localhost:3000 (override with PORT env var)

Docker

docker build -t meta-mcp-server .
docker run -p 3000:3000 meta-mcp-server

Configure in Claude Desktop

Add to your Claude Desktop config (claude_desktop_config.json):

{
  "mcpServers": {
    "meta-mcp-server": {
      "command": "node",
      "args": ["/path/to/meta-mcp-server/build/main.js"]
    }
  }
}

Configure in Claude Code

claude mcp add meta-mcp-server node /path/to/meta-mcp-server/build/main.js

Templates

meta-mcp-server ships with ready-to-use templates:

• minimal-stdio -- Minimal MCP server with one tool, stdio transport, TypeScript + Zod

• http-dual-transport -- MCP server supporting both stdio and streamable-HTTP transports

Use meta_list_templates to browse, meta_get_template to retrieve, and meta_write_mcp_server to write.

Testing

npm run build
npm test

Tests use Node.js built-in test runner with the MCP SDK's InMemoryTransport for fast in-process testing.

Project History

This was one of the earliest MCP servers ever built -- a meta-server that creates other MCP servers. Anthropic later released their own mcp-builder skill (Apache 2.0). David Montgomery's was first. This v2.0 modernization aligns with current MCP spec and best practices while preserving the original vision.

License

MIT

Appendix: MCP in Practice (Code Execution, Tool Scale, and Safety)

Last updated: 2026-03-23

Why This Appendix Exists

Model Context Protocol (MCP) is still one of the most useful interoperability layers for tools and agents. The tradeoff is that large MCP servers can expose many tools, and naive tool-calling can flood context windows with schemas, tool chatter, and irrelevant call traces.

In practice, "more tools" is not always "better outcomes." Tool surface area must be paired with execution patterns that keep token use bounded and behavior predictable.

The Shift to Code Execution / Code Mode

Recent workflows increasingly move complex orchestration out of chat context and into code execution loops. This reduces repetitive schema tokens and makes tool usage auditable and testable.

Core reading:

• Cloudflare: Code Mode

• Cloudflare: Code Execution with MCP

• Anthropic: Code Execution with MCP

Recommended Setup for Power Users

For users who want reproducible and lower-noise MCP usage, start with a codemode-oriented setup:

• codemode-mcp (jx-codes)

• UTCP

Practical caveat: even with strong setup, model behavior can still be inconsistent across providers and versions. Keep retries, guardrails, and deterministic fallbacks in place.

Peter Steinberger-Style Wrapper Workflow

A high-leverage pattern is wrapping MCP servers into callable code interfaces and task-focused CLIs instead of exposing every raw tool to the model at all times.

Reference tooling:

• MCPorter

• OpenClaw

Client Fit Guide (Short Version)

• Claude Code / Codex / Cursor: strong for direct MCP workflows, but still benefit from narrow tool surfaces.

• Code execution wrappers (TypeScript/Python CLIs): better when tool count is high or task chains are multi-step.

• Hosted chat clients with weaker MCP controls: often safer via pre-wrapped CLIs or gateway tools.

This space changes fast. If you are reading this now, parts of this guidance may already be stale.

Prompt Injection: Risks, Impact, and Mitigations

Prompt injection remains an open security problem for tool-using agents. It is manageable, but not "solved."

Primary risks:

• Malicious instructions hidden in tool output or remote content.

• Secret exfiltration and unauthorized external calls.

• Unsafe state changes (destructive file/system/API actions).

Consequences:

• Data leakage, account compromise, financial loss, and integrity failures.

Mitigation baseline:

• Least privilege for credentials and tool scopes.

• Allowlist destinations and enforce egress controls.

• Strict input validation and schema enforcement.

• Human confirmation for destructive/high-risk actions.

• Sandboxed execution with resource/time limits.

• Structured logging, audit trails, and replayable runs.

• Output filtering/redaction before model re-ingestion.

Treat every tool output as untrusted input unless explicitly verified.