home / mcp / legends mcp server

Legends MCP Server

Provides multi-legend conversations with Claude by running bundled legends via MCP, enabling diverse perspectives without external APIs.

Installation

Add the following to your MCP client configuration file.

Configuration

View docs

{
  "mcpServers": {
    "aytuncyildizli-legends-mcp": {
      "command": "npx",
      "args": [
        "legends-mcp"
      ]
    }
  }
}

You can run Legends MCP to enable Claude to chat with multiple legendary figures in a single conversation. This MCP server lets you activate multi-legend discussions, receive proactive suggestions, and automatically match legends to your topics, all without needing external APIs or keys.

How to use

Set up and interact with the Legends MCP to start conversations among multiple legends, get targeted recommendations, and let the system auto-match the best legends for your questions.

How to install

Prerequisites: ensure you have Node.js and a compatible runtime installed on your system. You will also need access to a client or environment that can consume MCP endpoints (such as Claude Code in this context). You will configure the MCP server as shown in the code snippet below.

{
  "mcpServers": {
    "legends": {
      "command": "npx",
      "args": ["legends-mcp"]
    }
  }
}

Additional sections

Security and usage features are included to protect prompts and sanitize inputs. The MCP server loads bundled legends by default, with optional opt-in for custom legends, and does not make external API calls. You can restart the Claude Code client after configuring the MCP server to begin chatting.

Environment and runtime details: the configuration uses a stdio style MCP server, executing a local command via npx to run legends-mcp. No API keys are required, and the setup relies on the bundled legend data.

You can use features like Party Mode for multi-legend discussions and Smart Suggest for proactive recommendations. The system can also auto-match legends to your topic to provide diverse perspectives.

Available tools

suggest

Analyzes your question and returns ready-to-execute actions to optimize legend usage and match the best legends for your topic.

party_mode

Coordinate and run multi-legend discussions, bringing 2 to 5 legends into a single conversation.

summon_legend

Channel the perspective of a single legend on a given topic.

auto_match

Deeply analyzes questions to identify which legends fit best for your topic.

list_legends

Browse all bundled legends available in the MCP package.

search_legends

Find legends by keyword, topic, or expertise.

get_legend_context

Provide a deep dive into a legend’s frameworks, principles, and examples.

get_legend_insight

Return a concise insight or takeaway from a legend on any topic.