Polymarket MCP Server
Enables comprehensive read access to Polymarket prediction market data, including live prices, order books, and market statistics. It allows AI assistants to analyze events, search markets, and monitor real-time signals without requiring API keys or authentication.
README
<p align="center"> <img src="https://polymarket.com/icons/favicon-196x196.png" width="80" height="80" alt="Polymarket logo" /> </p>
<h1 align="center">polymarket-mcp</h1>
<p align="center"> <strong>A production-grade MCP server for <a href="https://polymarket.com">Polymarket</a> prediction markets.</strong><br/> Give Claude (or any MCP client) full read access to every public Polymarket API. </p>
<p align="center"> <a href="#quick-start"><img src="https://img.shields.io/badge/Get_Started-blue?style=for-the-badge" alt="Get Started" /></a> <a href="https://www.npmjs.com/package/@yigit/polymarket-mcp"><img src="https://img.shields.io/npm/v/@yigit/polymarket-mcp?style=for-the-badge&logo=npm&color=cb3837" alt="npm" /></a> <a href="LICENSE"><img src="https://img.shields.io/badge/license-MIT-green?style=for-the-badge" alt="MIT License" /></a> <a href="https://modelcontextprotocol.io"><img src="https://img.shields.io/badge/MCP-compatible-8A2BE2?style=for-the-badge" alt="MCP Compatible" /></a> </p>
Why?
Prediction markets are the best real-time signal for what the world thinks will happen next. This server puts that signal directly into your AI assistant's toolkit — no API keys, no auth tokens, no setup hassle.
23 tools • 5 resources • 4 prompts • Zero config • Zero dependencies beyond the MCP SDK
Quick Start
Option 1: Claude Desktop
Add to your claude_desktop_config.json (Settings > Developer > Edit Config):
{
"mcpServers": {
"polymarket": {
"command": "npx",
"args": ["-y", "@yigit/polymarket-mcp"]
}
}
}
Restart Claude Desktop. Done.
Option 2: Claude Code
claude mcp add polymarket -- npx -y @yigit/polymarket-mcp
Option 3: Build from source
git clone https://github.com/yigitabi5444/yigit_polymarket_mcp.git
cd yigit_polymarket_mcp
npm install && npm run build
node dist/index.js # runs over stdio
Then point your MCP client at the built binary:
{
"mcpServers": {
"polymarket": {
"command": "node",
"args": ["/absolute/path/to/yigit_polymarket_mcp/dist/index.js"]
}
}
}
What Can It Do?
Gamma API — Events & Markets
| Tool | What it does |
|---|---|
get_events |
List/filter events — paginate, sort by volume/liquidity, filter by tag/status |
get_event |
Single event by ID or slug |
get_markets |
List/filter markets with rich filtering (volume, liquidity, dates, tags) |
get_market |
Single market by ID or slug |
search |
Full-text search across events, markets, and profiles |
get_tags |
List all category tags |
get_series |
List event series (grouped collections) |
get_series_by_id |
Get a specific series by ID |
get_sports |
List available sports |
get_sports_teams |
List teams for a sport |
CLOB API — Prices & Order Books
| Tool | What it does |
|---|---|
get_price |
Current price for a token (buy/sell side) |
get_midpoint |
Midpoint price (best bid + best ask / 2) |
get_last_trade_price |
Last executed trade price |
get_price_history |
Historical price time series (configurable interval + fidelity) |
get_order_book |
Full bid/ask depth for a token |
get_order_books |
Batch order books for multiple tokens |
get_order_book_summary |
Best bid, best ask, and spread at a glance |
get_clob_market |
CLOB market details by condition ID |
get_sampling_markets |
Markets eligible for liquidity rewards |
get_sampling_simplified_markets |
Simplified sampled markets |
get_tick_size |
Minimum price increment for a token |
Data API — Trades & Holders
| Tool | What it does |
|---|---|
get_market_trades |
Recent trades for a market |
get_market_holders |
Top holders / position breakdown for a market |
Resources (URI-based access)
| URI Pattern | Description |
|---|---|
market://slug/{slug} |
Market JSON by slug |
event://slug/{slug} |
Event JSON by slug |
market://condition/{id} |
CLOB market by condition ID |
orderbook://token/{id} |
Live order book for a token |
tags://all |
All category tags (cached) |
Prompts (guided workflows)
| Prompt | Description |
|---|---|
analyze_market |
Deep-dive: probability, liquidity, order book depth, top holders |
compare_markets |
Side-by-side comparison of multiple markets |
trending_markets |
Find the hottest markets by volume right now |
sports_overview |
Overview of sports prediction markets |
Architecture
┌─────────────────────────────┐
│ MCP Client │
│ (Claude Desktop / Code) │
└─────────┬───────────────────┘
│ stdio (JSON-RPC)
┌─────────▼───────────────────┐
│ polymarket-mcp │
│ │
│ ┌──────────┐ ┌──────────┐ │
│ │ Tools │ │Resources │ │
│ │ (22) │ │ (5) │ │
│ └────┬─────┘ └────┬─────┘ │
│ │ │ │
│ ┌────▼─────────────▼─────┐ │
│ │ API Client │ │
│ │ ┌───────┐ ┌────────┐ │ │
│ │ │ Rate │ │ Cache │ │ │
│ │ │Limiter│ │ (TTL) │ │ │
│ │ └───┬───┘ └───┬────┘ │ │
│ │ │ │ │ │
│ │ ┌───▼─────────▼────┐ │ │
│ │ │ Retry + Backoff │ │ │
│ │ └──────────────────┘ │ │
│ └────────────────────────┘ │
└──────────┬───────────────────┘
│ HTTPS
┌──────────▼───────────────────┐
│ Polymarket APIs │
│ gamma-api · clob · data-api │
└──────────────────────────────┘
Smart Defaults
List endpoints (get_markets, get_events) default to active=true, closed=false, order=volume descending — you get the hottest live markets out of the box, not ancient resolved ones from 2020. Override any default explicitly when you need historical data.
Clean Responses
Raw Polymarket API responses contain 50+ fields per market (including gems like pagerDutyNotificationEnabled and mailchimpTag). This server strips the noise and returns a curated subset. JSON-encoded strings like outcomes and outcomePrices are parsed into real arrays — ["Yes", "No"] not "[\"Yes\", \"No\"]".
Production Hardening
| Feature | Implementation |
|---|---|
| Rate limiting | Token-bucket per API endpoint (Gamma: 4k/10s, CLOB: 9k/10s, Data: 1k/10s) |
| Caching | In-memory TTL cache (order books: 5s, markets: 30s, tags: 5min) |
| Retries | Exponential backoff with jitter on 429s and 5xx errors (3 retries max) |
| Timeouts | 15s request timeout on all API calls |
| No auth | Only hits public endpoints — zero secrets to manage |
| No deps | Uses native fetch — no axios, no node-fetch, just the MCP SDK + Zod |
| Type safety | Full TypeScript with strict mode, Zod validation on all inputs |
Configuration
| Environment Variable | Default | Description |
|---|---|---|
POLYMARKET_CACHE_DISABLED |
false |
Set to true to disable all caching |
That's it. No API keys, no tokens, no config files. It just works.
Development
npm install # install dependencies
npm run build # compile TypeScript
npm run dev # run with tsx (auto-reload)
npm test # run test suite
npm run test:watch # tests in watch mode
Project Structure
src/
├── index.ts # stdio entrypoint
├── server.ts # MCP server factory
├── config.ts # rate limits, cache TTLs, timeouts
├── api/
│ ├── client.ts # HTTP client with retry + rate limiting
│ ├── cache.ts # TTL cache implementation
│ ├── rate-limiter.ts # token-bucket rate limiter
│ ├── gamma.ts # Gamma API wrapper
│ ├── clob.ts # CLOB API wrapper
│ └── data.ts # Data API wrapper
├── tools/
│ ├── gamma/ # 6 Gamma tools
│ ├── clob/ # 4 CLOB tools
│ └── data/ # 2 Data tools
├── resources/ # 5 URI-based resources
├── prompts/ # 4 guided analysis prompts
└── types/ # TypeScript type definitions
Example Conversations
You: What are the hottest prediction markets right now?
Claude: uses
get_marketssorted by volume, thenget_order_bookfor depth analysis
You: What's the current probability that Bitcoin hits $200k this year?
Claude: uses
search("bitcoin 200k"), thenget_pricefor live odds
You: Compare the presidential election markets
Claude: uses the
compare_marketsprompt with relevant slugs
You: Show me the order book for the top sports market
Claude: uses
get_sports→get_markets→get_order_book
Contributing
PRs welcome! Please:
- Fork the repo
- Create a feature branch (
git checkout -b feature/awesome) - Add tests for new functionality
- Make sure
npm testpasses - Submit a PR
License
MIT — do whatever you want with it.
<p align="center"> <sub>Built with the <a href="https://modelcontextprotocol.io">Model Context Protocol</a></sub> </p>
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.