The Polymarket MCP Server provides an AI-powered trading assistant that enables Claude to autonomously monitor Polymarket markets, analyze data, and place trades within safety constraints. It exposes a suite of 45 tools across market discovery, market analysis, trading, portfolio management, and real-time monitoring, including AI-driven opportunity analysis, AI-suggested pricing, risk scoring, position rebalancing, live WebSocket feeds, and enterprise-grade safety checks such as order size limits, exposure caps, liquidity validation, and pre-trade validation. You can run the server locally and use the web dashboard for real-time status, configuration tuning, and market exploration, while Claude can leverage the 45 tools to make informed trading decisions.

Prerequisites:

• Python 3.10 or newer
• Git
• (Optional) Virtual environment tool (venv is included in Python)

Step-by-step installation:

1. Clone the repository: git clone https://github.com/caiovicentino/polymarket-mcp-server.git cd polymarket-mcp-server

2. Create and activate a virtual environment: python -m venv venv

   macOS/Linux

   source venv/bin/activate

   Windows

   venv\Scripts\activate

3. Install the package in editable mode: pip install -e .

4. Configure required credentials (demo vs full mode):

   • For DEMO mode: cp .env.example .env and set DEMO_MODE=true
   • For Full Trading mode: copy and edit .env with wallet credentials: POLYGON_PRIVATE_KEY=your_private_key_without_0x_prefix POLYGON_ADDRESS=0xYourPolygonAddress

5. Install optional dependencies and start the server as described in the README (manual approach):

   Quick start (demo environment)

   curl -sSL https://raw.githubusercontent.com/caiovicentino/polymarket-mcp-server/main/quickstart.sh | bash

6. Optional Claude Desktop integration (for Claude workflow):

   • Add an mcpServers entry in Claude Desktop config as shown in the README

Notes:

• The repository provides quickstart.sh for a one-command setup and an install.sh script for a full trading setup. Follow the preferred path depending on whether you want a quick demo or a production-like environment.

Tips and common considerations:

• Ensure your POLYGON_PRIVATE_KEY is kept secret and never exposed in public repos.
• Adjust safety parameters in your .env file to fit your risk tolerance: MAX_ORDER_SIZE_USD, MAX_TOTAL_EXPOSURE_USD, MAX_POSITION_SIZE_PER_MARKET, MIN_LIQUIDITY_REQUIRED, MAX_SPREAD_TOLERANCE, ENABLE_AUTONOMOUS_TRADING, and REQUIRE_CONFIRMATION_ABOVE_USD.
• The web dashboard is available at http://localhost:8080 and provides real-time status, configuration sliders for safety limits, and market analyses.
• If you encounter WebSocket disconnects, enable auto-reconnect with exponential backoff and ensure network stability.
• For Claude Desktop integration, mirror the mcpServers structure shown in the README to enable seamless cross-application trading workflows.