Polymarket MCP Server
Provides AI-powered tools for searching Polymarket prediction markets and calculating financial metrics like expected value and Kelly bet sizing. It enables users to perform arbitrage scanning, news-based sentiment analysis, and market research through natural language interfaces.
README
Polymarket MCP -- The Open-Source Prediction Market Agent
Your AI edge in prediction markets. Search, analyze, detect mispricing, track whales, and trade -- all from your LLM.
What This Is
Most prediction market tools give you a dashboard. This gives you a brain. Polymarket MCP connects any LLM (Claude, GPT, Gemini) to Polymarket's data and wraps it with EV calculations, Kelly sizing, arbitrage detection, news-driven edge finding, and crypto correlation analysis -- all as MCP tools your AI can call autonomously.
Not a toy. Not a wrapper. A full agent toolkit for prediction markets.
Features
| Category | Capability | Status |
|---|---|---|
| Market Analysis | Search, trending, details, sports/crypto filters, market summaries | Live (9 tools) |
| Edge Detection | Auto-scan markets for mispricing via news sentiment analysis | Live (2 tools) |
| Risk Management | EV calculation, Kelly criterion sizing, arbitrage scanning | Live (3 tools) |
| Crypto Integration | Live prices (CoinGecko), Fear/Greed index, crypto vs predictions | Live (4 tools) |
| Paper Trading | Risk-free practice with virtual bankroll, track P&L | Planned |
| Whale Tracking | Follow top traders (Theo4, Fredi9999, etc.), copy signals | Planned |
| Alert System | Telegram/Discord notifications on edge detection triggers | Planned |
| Live Trading | Execute trades via Polymarket CLOB API | Planned |
Quick Start
git clone https://github.com/Miles0sage/polymarket-mcp.git
cd polymarket-mcp
pip install -r requirements.txt
Add to your MCP config (~/.mcp.json or client settings):
{
"polymarket": {
"command": "python3",
"args": ["server.py"],
"cwd": "/path/to/polymarket-mcp"
}
}
No API keys required for market data. Works immediately with Claude Code, Cursor, Windsurf, or any MCP-compatible client.
Architecture
+------------------+
| LLM / Client |
| (Claude, GPT...) |
+--------+---------+
|
MCP Protocol
|
+--------v---------+
| MCP Server |
| (server.py) |
+--------+---------+
|
+-------------------+-------------------+
| | |
+--------v------+ +--------v------+ +---------v-----+
| Market Data | | Analysis | | Research |
| (tools_ | | (tools_ | | (tools_ |
| markets.py) | | analysis.py)| | research.py) |
+-------+-------+ +-------+-------+ +--------+------+
| | |
+--------v------+ +-------v-------+ +---------v-----+
| Gamma API | | EV / Kelly / | | Google News |
| CLOB API | | Arbitrage | | YouTube |
+---------------+ +---------------+ | Sentiment |
+---------------+
+-------------------+
| Crypto Data |
| (tools_crypto.py)|
+--------+----------+
|
+--------v----------+
| CoinGecko API |
| Fear/Greed Index |
+---------+---------+
|
(Future)
|
+------------------v------------------+
| Paper Trading | Whale Tracker |
| Risk Manager | Alert System |
| CLOB Trading | Dashboard UI |
+------------------+------------------+
Tool Reference
Market Data (5 tools)
| Tool | Description |
|---|---|
search_markets |
Search active markets by keyword |
get_market |
Full details for a specific market (prices, volume, liquidity, dates) |
trending_markets |
Top markets ranked by trading volume |
crypto_prediction_markets |
Active crypto-related markets (BTC, ETH, SOL, DeFi) |
sports_prediction_markets |
Active sports markets (NBA, NFL, UFC, F1, NCAA) |
Analysis (3 tools)
| Tool | Description |
|---|---|
calculate_ev |
Expected value + Kelly sizing for YES/NO positions |
kelly_size |
Full and quarter Kelly bet sizing given bankroll, probability, odds |
arbitrage_scan |
Detect arbitrage across market outcomes |
Research & Edge Finding (3 tools)
| Tool | Description |
|---|---|
research_market |
News + YouTube + sentiment analysis for any market question |
edge_finder |
Batch scan markets for mispricing via news sentiment vs market price |
market_summary |
Human-readable summary with implied probability and market read |
Crypto (4 tools)
| Tool | Description |
|---|---|
crypto_prices |
Live prices from CoinGecko with 24h change, volume, market cap |
crypto_overview |
Top 20 crypto by market cap |
crypto_fear_greed |
Bitcoin Fear & Greed Index (7-day history + trading signal) |
crypto_vs_predictions |
Compare spot price with related Polymarket markets |
Usage Examples
"Search Polymarket for Bitcoin markets"
-> 10 markets with YES/NO prices, volume, liquidity
"Calculate EV: YES price 0.35, my estimate 0.55"
-> EV: +57.1%, Quarter-Kelly: 7.7% of bankroll, BET YES
"Check for arbitrage: 0.45, 0.43, 0.08"
-> ARBITRAGE FOUND! Buy all for $0.96, collect $1.00
"Research: Will Bitcoin hit $200K by end of 2026?"
-> News headlines, YouTube analysis, sentiment score
"Compare BTC spot price with Polymarket prediction markets"
-> Current price + all related markets side by side
"What's the Fear & Greed Index saying?"
-> 7-day chart, current reading, trading signal
Roadmap
- [ ] Paper Trading -- Virtual bankroll, position tracking, P&L history
- [ ] Whale Tracker -- Monitor top Polymarket wallets, alert on large positions
- [ ] Live Trading -- Execute via Polymarket CLOB API (requires wallet + API key)
- [ ] WebSocket Feeds -- Real-time price updates instead of polling
- [ ] Alert System -- Telegram/Discord notifications on edge triggers
- [ ] Dashboard UI -- Web UI for portfolio, positions, edge scanner
- [ ] Discord Bot -- Query markets and get alerts in Discord
- [ ] One-Click Deploy -- Railway/Render template for hosted instance
Comparison
| Feature | Polymarket MCP | Polystrat | Kalshi News Bot | Manual Trading |
|---|---|---|---|---|
| LLM integration | Native MCP | None | API only | None |
| Edge detection | Auto (news sentiment) | Manual | None | Manual |
| Kelly sizing | Built-in | None | None | Spreadsheet |
| Arbitrage scan | Built-in | None | None | Manual |
| Crypto correlation | Live CoinGecko | None | N/A | Manual |
| Open source | MIT | Closed | Closed | N/A |
| Setup time | 2 minutes | Account required | API key + setup | N/A |
| Cost | Free | Subscription | Subscription | Free |
Contributing
PRs welcome. The codebase is intentionally simple -- pure Python, no frameworks, no databases. Each tool module is standalone.
server.py # MCP server, tool registration
tools_markets.py # Market data (Gamma API, CLOB API)
tools_analysis.py # EV, Kelly, arbitrage math
tools_research.py # News sentiment, edge finding
tools_crypto.py # CoinGecko prices, Fear/Greed
License
MIT
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.