brainstorm-mcp

An MCP server that runs multi-round brainstorming debates between AI models. Connect it to Claude Code (or any MCP client) and let GPT, DeepSeek, Groq, Ollama, and others debate your ideas — then get a synthesized final output.

How it works

1. You ask Claude: "Brainstorm the best architecture for a real-time app"
2. The tool sends the topic to all configured AI models in parallel
3. Each model responds independently (Round 1)
4. Models see each other's responses and refine their positions (Rounds 2-N)
5. A synthesizer model produces a final consolidated output
6. You get back a structured debate with the synthesis

Quick Start

# Clone and build
git clone https://github.com/AIPoweredSolutions/brainstorm-mcp.git
cd brainstorm-mcp
npm install
npm run build

Configure providers

Copy the example config and add your API keys:

cp brainstorm.config.example.json brainstorm.config.json

Edit brainstorm.config.json:

{
  "providers": {
    "openai": {
      "model": "gpt-4o",
      "apiKeyEnv": "OPENAI_API_KEY"
    },
    "deepseek": {
      "model": "deepseek-chat",
      "apiKeyEnv": "DEEPSEEK_API_KEY"
    }
  }
}

Connect to Claude Code

Add to your project's .mcp.json:

{
  "mcpServers": {
    "brainstorm": {
      "command": "node",
      "args": ["/path/to/brainstorm-mcp/dist/index.js"],
      "env": {
        "OPENAI_API_KEY": "sk-...",
        "DEEPSEEK_API_KEY": "sk-...",
        "BRAINSTORM_CONFIG": "/path/to/brainstorm.config.json"
      }
    }
  }
}

Restart Claude Code, then just ask:

"Brainstorm the best way to handle authentication in a microservices architecture"

Configuration

brainstorm.config.json

The config file defines AI providers. Known providers (openai, deepseek, groq, mistral, together) don't need a baseURL — it's auto-detected.

{
  "providers": {
    "openai": {
      "model": "gpt-4o",
      "apiKeyEnv": "OPENAI_API_KEY"
    },
    "deepseek": {
      "model": "deepseek-chat",
      "apiKeyEnv": "DEEPSEEK_API_KEY"
    },
    "groq": {
      "model": "llama-3.3-70b-versatile",
      "apiKeyEnv": "GROQ_API_KEY"
    },
    "ollama": {
      "model": "llama3.1",
      "baseURL": "http://localhost:11434/v1"
    }
  }
}

Fallback: Environment Variables

If no config file exists, the server detects providers from env vars:

OPENAI_API_KEY=sk-...
OPENAI_DEFAULT_MODEL=gpt-4o
DEEPSEEK_API_KEY=sk-...
DEEPSEEK_DEFAULT_MODEL=deepseek-chat

MCP Tools

brainstorm

Run a multi-round debate. Only topic is required — everything else has sensible defaults.

list_providers

Shows all configured providers, their default models, and API key status.

add_provider

Dynamically add a provider at runtime.

Features

• Multi-round debates — Models see and critique each other's responses
• Parallel execution — All models respond concurrently within each round
• Per-model timeouts — 2-minute timeout per API call, one slow model won't block others
• Context truncation — Automatically truncates history when approaching context limits
• Cost estimation — Shows estimated token usage and cost
• Resilient — One model failing doesn't abort the debate
• Synthesizer fallback — If the primary synthesizer fails, tries other models
• GPT-5.x / o3 / o4 compatible — Automatically uses max_completion_tokens for newer OpenAI models

License

MIT