Multi-Agent Architecture Knowledge MCP

An MCP server that provides expert knowledge on multi-agent architectures directly to your AI coding assistants (Claude Code, Cursor, Claude Desktop).

What does this do?

When building multi-agent systems, your AI coding assistant can instantly access:

• Architecture Patterns: Supervisor/Worker, Hierarchical, Swarm, Blackboard, Pipeline, Contract-Net
• Framework Recommendations: LangGraph, CrewAI, AutoGen, MetaGPT, AgentScope
• Implementation Strategies: Step-by-step guidance with code examples
• Use Case Analysis: Pattern recommendations tailored to your requirements
• Bilingual Support: English and Japanese queries

How this helps: Instead of spending hours researching which architecture pattern fits your use case, your AI agent accesses curated knowledge from arXiv papers and official documentation to design better multi-agent systems in real-time.

Setup Instructions

No local installation needed - just add the URL to your MCP client configuration.

Claude Code

claude mcp add multi-agent-architect \
  --url https://ai-crew-solutions--multi-agent-architect-mcp.apify.actor \
  --token YOUR_APIFY_API_TOKEN

Get your Apify API token from: https://console.apify.com/account/integrations

Cursor

1. Open your Cursor MCP configuration file:

   • Global: ~/.cursor/mcp.json
   • Project-specific: .cursor/mcp.json in your project root

2. Add this configuration:

{
  "mcpServers": {
    "multi-agent-architect": {
      "url": "https://ai-crew-solutions--multi-agent-architect-mcp.apify.actor",
      "headers": {
        "Authorization": "Bearer YOUR_APIFY_API_TOKEN"
      }
    }
  }
}

3. Restart Cursor

Claude Desktop

1. Open your Claude Desktop configuration file:

   • macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
   • Windows: %APPDATA%\Claude\claude_desktop_config.json

2. Add this configuration:

{
  "mcpServers": {
    "multi-agent-architect": {
      "url": "https://ai-crew-solutions--multi-agent-architect-mcp.apify.actor",
      "headers": {
        "Authorization": "Bearer YOUR_APIFY_API_TOKEN"
      }
    }
  }
}

3. Restart Claude Desktop

Available Tools

search_knowledge

Search the multi-agent architecture knowledge base.

Parameters:

• query (string, required): Your question in English or Japanese
• top_k (number, optional): Number of results (default: 5)

Example:

Ask your AI agent: "What is the Supervisor/Worker pattern and when should I use it?"

recommend_architecture

Get architecture pattern recommendations for your specific use case.

Parameters:

• use_case (string, required): Description of what you're building
• constraints (object, optional): Requirements like complexity, scalability

Example:

Ask your AI agent: "Recommend an architecture for a large-scale customer support automation system"

Example Usage

Once configured, simply ask your AI coding assistant questions like:

• "I'm building a customer support chatbot with escalation. Which multi-agent pattern should I use?"
• "Compare LangGraph vs CrewAI for building a research assistant with multiple specialized agents"
• "What's the best architecture for parallel task processing with 10+ agents?"

Your AI agent will use this MCP server to provide expert recommendations backed by research papers and production case studies.

Knowledge Base

• Architecture Patterns: 6+ major multi-agent patterns with use case mappings
• Framework Comparisons: LangGraph, CrewAI, AutoGen, MetaGPT, AgentScope
• Use Case Library: 20+ real-world scenarios with pattern recommendations
• Sources: Curated from arXiv papers, official documentation, and production case studies

Pricing

• search_knowledge: $0.30 per query
• recommend_architecture: $0.40 per recommendation

Why? Saves 3-5 hours of research time per project. Pay only when your AI agent actually uses the tools.

Version: 1.0.0 Built by: AI Crew Solutions