Meta Prompt MCP
A server that transforms a standard Language Model into a dynamic multi-agent system where the model simulates both a Conductor (project manager) and Experts (specialized agents) to tackle complex problems through a collaborative workflow.
What is Meta Prompt MCP?
Meta Prompt MCP is a Model Context Protocol (MCP) server that allows AI assistants like Claude, Cursor, and VS Code to server that transforms a standard language model into a dynamic multi-agent system where the model simulates both a conductor (project manager) and experts (specialized agents) to tackle complex probl...
A server that transforms a standard Language Model into a dynamic multi-agent system where the model simulates both a Conductor (project manager) and Experts (specialized agents) to tackle complex problems through a collaborative workflow.
This server falls under the Coding Agents category on MCPgee, the world's largest MCP server directory with 33,000+ servers.
Features
- A server that transforms a standard Language Model into a dy
Use Cases
Maintainer
Works with
Installation
Manual Installation
npx meta-prompt-mcp-serverConfiguration
Configuration Details
claude_desktop_config.json
Performance
Response Metrics
Resource Usage
How to Set Up and Use Meta Prompt MCP
Meta Prompt MCP Server implements the Meta-Prompting technique from the Stanford/OpenAI research paper 'Meta-Prompting: Enhancing Language Models with Task-Agnostic Scaffolding'. It transforms a single LLM into a dynamic multi-agent system by having the model alternate between two roles: a Conductor (high-level project manager that breaks complex problems into subtasks and delegates) and Experts (specialized agents such as 'Python Programmer' or 'Code Reviewer' consulted to solve each subtask). This scaffolding leads to more thorough, self-reviewed solutions without requiring separate model instances.
Prerequisites
- Python 3.10 or higher
- uv package manager installed (https://astral.sh/uv) — the project requires uv to run
- The meta-prompt-mcp-server repository cloned locally
- An MCP-compatible client such as Claude Desktop or Cursor
Clone the repository
Download the meta-prompt-mcp-server source code to your local machine.
git clone https://github.com/tisu19021997/meta-prompt-mcp-server.git
cd meta-prompt-mcp-serverInstall uv
This project requires uv, the fast Python package manager from Astral. Install it for your platform. After installation, verify with `uv --version` and note the full path returned by `which uv` — you will need it in the config.
# macOS / Linux:
curl -LsSf https://astral.sh/uv/install.sh | sh
# Windows PowerShell:
powershell -c "irm https://astral.sh/uv/install.ps1 | iex"
# Verify and get path:
which uvConfigure Claude Desktop
Add the server entry to claude_desktop_config.json (macOS: ~/Library/Application Support/Claude/claude_desktop_config.json). Replace the placeholder paths with the actual absolute paths for uv and the cloned repository on your machine.
{
"mcpServers": {
"meta-prompting": {
"command": "/path/to/your/uv",
"args": [
"--directory",
"/path/to/your/meta-prompt-mcp-server",
"run",
"mcp-meta-prompt"
]
}
}
}Restart Claude Desktop
Fully quit and relaunch Claude Desktop so it picks up the new server configuration and the meta-prompting server registers its tools.
Invoke the meta_model_prompt tool to activate the workflow
In your conversation, explicitly invoke the `meta_model_prompt` tool from the `meta-prompting` server. This is the entry point that activates the Conductor/Expert workflow. Then provide your complex problem statement as the query.
Meta Prompt MCP Examples
Client configuration (Claude Desktop)
Complete claude_desktop_config.json entry for the Meta Prompt MCP server. You must use absolute paths to both uv and the cloned repository.
{
"mcpServers": {
"meta-prompting": {
"command": "/Users/you/.local/bin/uv",
"args": [
"--directory",
"/Users/you/meta-prompt-mcp-server",
"run",
"mcp-meta-prompt"
]
}
}
}Prompts to try
How to use the Meta Prompt MCP server — always start by invoking the meta_model_prompt tool, then provide your problem.
- Invoke meta_model_prompt with query: "Design a REST API for a task management system with authentication, rate limiting, and proper error handling."
- Invoke meta_model_prompt with query: "Write and review a Python function that parses complex nested JSON and handles all edge cases."
- Invoke meta_model_prompt with query: "Plan and outline a technical blog post explaining how WebSockets differ from HTTP polling."Troubleshooting Meta Prompt MCP
Server fails to start with 'command not found' for uv
Use the full absolute path to the uv binary in the command field (e.g., /Users/you/.local/bin/uv or /home/you/.cargo/bin/uv). Run `which uv` in your terminal to get the exact path, then update your MCP config.
meta_model_prompt tool is not visible or the expert_model tool has no effect
Most MCP clients (including current versions of Claude Desktop and Cursor) do not yet implement ctx.sample(), which is needed for true independent expert model calls. The server falls back to including the expert output within the conductor's single generation — this is expected behavior, not a bug. The workflow still produces better results than a plain prompt.
uv cannot find the project or throws a 'no pyproject.toml' error
Ensure the --directory argument in your args array points to the exact root of the cloned meta-prompt-mcp-server repository (the folder containing pyproject.toml). Use an absolute path, not a relative one.
Frequently Asked Questions about Meta Prompt MCP
What is Meta Prompt MCP?
Meta Prompt MCP is a Model Context Protocol (MCP) server that server that transforms a standard language model into a dynamic multi-agent system where the model simulates both a conductor (project manager) and experts (specialized agents) to tackle complex problems through a collaborative workflow. It connects AI assistants to external tools and data sources through a standardized interface.
How do I install Meta Prompt MCP?
Follow the installation instructions on the Meta Prompt MCP GitHub repository. Clone the repo, install dependencies, and add the server config to your AI client.
Which AI clients work with Meta Prompt MCP?
Meta Prompt MCP works with all major MCP-compatible AI clients including Claude Desktop, Claude Code, Cursor, VS Code (GitHub Copilot), Windsurf, and Cline.
Is Meta Prompt MCP free to use?
Yes, Meta Prompt MCP is open source and available under the Apache 2.0 license. You can use it freely in both personal and commercial projects.
Meta Prompt MCP Alternatives — Similar Coding Agents Servers
Looking for alternatives to Meta Prompt MCP? Here are other popular coding agents servers you can use with Claude, Cursor, and VS Code.
Dify
★ 142.2kProduction-ready platform for agentic workflow development.
Ruflo
★ 54.0k🌊 The leading agent orchestration platform for Claude. Deploy intelligent multi-agent swarms, coordinate autonomous workflows, and build conversational AI systems. Features enterprise-grade architecture, self-learning swarm intelligence, RAG integrat
Goose
★ 45.7kan open source, extensible AI agent that goes beyond code suggestions - install, execute, edit, and test with any LLM
Antigravity Awesome Skills
★ 38.3kInstallable GitHub library of 1,400+ agentic skills for Claude Code, Cursor, Codex CLI, Gemini CLI, Antigravity, and more. Includes installer CLI, bundles, workflows, and official/community skill collections.
AgentScope
★ 25.5kBuild and run agents you can see, understand and trust.
Serena
★ 24.5kA coding agent toolkit that provides IDE-like semantic code retrieval and editing tools, enabling LLMs to efficiently navigate and modify codebases using symbol-level operations instead of basic file reading and string replacements.
Browse More Coding Agents MCP Servers
Explore all coding agents servers available in the MCPgee directory. Each server includes setup guides for Claude, Cursor, and VS Code.
Set Up Meta Prompt MCP in Your Editor
Choose your AI client for step-by-step setup instructions.
Quick Config Preview
Add this to your claude_desktop_config.json or .cursor/mcp.json
Ready to use Meta Prompt MCP?
Browse our complete directory of 33,000+ MCP servers, read setup guides for your editor, and start building with the Model Context Protocol.