A Model Context Protocol (MCP) server that exposes AI-powered investor persona simulations through the MCP protocol.
Features
Individual Investor Personas
- Fred Wilson (Union Square Ventures): Simulation of the NYC-based VC known for community-driven ventures
- Peter Thiel (Founders Fund): Simulation of the contrarian investor focused on disruptive technologies
Multi-Persona Orchestrator
- Persona Orchestrator: Intelligent agent that routes questions to appropriate investor personas and explains their rationale
Installation
Prerequisites
- Python 3.11+
- uv package manager
- OpenAI API key
Install via Smithery
npx -y @smithery/cli install @lroolle/openai-agents-mcp-server --client claude
Claude Desktop Configuration
"mcpServers": {
"investor-persona-mcp-server": {
"command": "uvx",
"args": ["investor-persona-mcp-server"],
"env": {
"OPENAI_API_KEY": "your-api-key-here"
}
}
}
Implementation Details
Persona Configuration
Investor personas are defined through markdown files containing:
- Investment philosophy
- Psychological profile
- Historical track record
- Decision-making patterns
- Communication style preferences
Customization Options
- Add new investor personas by creating markdown profiles
- Implement custom interaction patterns between personas
- Enhance orchestration logic for complex multi-perspective analysis
Configuration
Environment variables:
- OPENAI_API_KEY: Required for AI model access
- MCP_TRANSPORT: Transport protocol (default: "stdio")
- PERSONAS_DIR: Path to investor profiles (default: "./investors")
Development
# Clone repo
git clone https://github.com/lroolle/investor-persona-mcp-server.git
cd investor-persona-mcp-server
# Setup environment
uv venv
source .venv/bin/activate
uv sync --dev
# Run with SSE transport
export OPENAI_API_KEY=your-key
uv run mcp dev src/agents_mcp_server/server.py
Test via MCP Inspector at http://localhost:5173
License
MIT