CoinGlass MCP Server
Provides AI agents access to cryptocurrency derivatives analytics through 22 unified tools covering 80+ CoinGlass API endpoints. Enables analysis of open interest, funding rates, liquidations, whale positions, order book data, and market sentiment indicators with plan-aware feature gating.
README
CoinGlass MCP Server
MCP server for CoinGlass cryptocurrency derivatives analytics. Provides AI agents access to 80+ API endpoints through 22 unified tools.
Features
- 22 MCP Tools — Unified interface to 80+ CoinGlass API endpoints
- Plan-Aware Gating — Automatic feature restrictions based on subscription tier
- Response Caching — Built-in caching via FastMCP middleware (60s TTL)
- Retry Logic — Automatic retries for transient failures (5xx, timeouts)
- Type-Safe — Full type hints with Literal-typed actions for LLM clarity
- Async-First — Built on httpx + FastMCP for high performance
Quick Start
Installation
pip install coinglass-mcp
Or with uv:
uv pip install coinglass-mcp
Configuration
export COINGLASS_API_KEY="your-api-key"
export COINGLASS_PLAN="standard" # hobbyist | startup | standard | professional | enterprise
Get your API key at coinglass.com/pricing
Run
coinglass-mcp
Claude Desktop Integration
Add to your claude_desktop_config.json:
{
"mcpServers": {
"coinglass": {
"command": "coinglass-mcp",
"env": {
"COINGLASS_API_KEY": "your-api-key",
"COINGLASS_PLAN": "standard"
}
}
}
}
Available Tools
| Category | Tool | Description |
|---|---|---|
| Market | coinglass_market_info |
Supported coins, pairs, exchanges |
coinglass_market_data |
Real-time market summaries | |
coinglass_price_history |
OHLC price candles | |
| Open Interest | coinglass_oi_history |
OI OHLC (pair/aggregated/stablecoin/coin-margin) |
coinglass_oi_distribution |
OI breakdown by exchange | |
| Funding | coinglass_funding_history |
Funding rate OHLC |
coinglass_funding_current |
Current rates, accumulated, arbitrage | |
| Long/Short | coinglass_long_short |
Global ratio, top accounts, top positions |
| Liquidation | coinglass_liq_history |
Liquidation OHLC history |
coinglass_liq_orders |
Real-time liquidation stream ⚡ | |
coinglass_liq_heatmap |
Liquidation heatmaps 🔥 | |
| Order Book | coinglass_ob_history |
Bid/ask depth history |
coinglass_ob_large_orders |
Whale walls detection | |
| Whale | coinglass_whale_positions |
Hyperliquid whale positions ⚡ |
coinglass_whale_index |
Whale activity index | |
| Taker | coinglass_taker |
Taker buy/sell volume and ratio |
| Spot | coinglass_spot |
Spot market data and prices |
| Options | coinglass_options |
Max pain, OI, volume (BTC/ETH) |
| On-Chain | coinglass_onchain |
Exchange balances, flows, transfers |
| ETF | coinglass_etf |
Bitcoin/Ethereum ETF flows |
coinglass_grayscale |
Grayscale holdings and premium | |
| Indicators | coinglass_indicators |
RSI, Fear & Greed, Rainbow, Pi Cycle, etc. |
| Meta | coinglass_search |
Discover tools by keyword |
coinglass_config |
View exchanges, intervals, features |
⚡ Requires Startup+ plan | 🔥 Requires Professional+ plan
Plan Features
| Feature | Hobbyist | Startup | Standard | Professional |
|---|---|---|---|---|
| Basic intervals (h4, h8, d1) | ✅ | ✅ | ✅ | ✅ |
| Extended intervals (m1-h1) | ❌ | ✅ | ✅ | ✅ |
| Whale alerts & positions | ❌ | ✅ | ✅ | ✅ |
| Liquidation orders stream | ❌ | ❌ | ✅ | ✅ |
| Liquidation heatmaps | ❌ | ❌ | ❌ | ✅ |
Usage Examples
Market Overview
# Get all coins summary
coinglass_market_data(action="coins_summary")
# Get BTC metrics only
coinglass_market_data(action="coins_summary", symbol="BTC")
Open Interest Analysis
# BTC OI across all exchanges
coinglass_oi_history(action="aggregated", symbol="BTC")
# OI distribution by exchange
coinglass_oi_distribution(action="by_exchange", symbol="BTC")
Funding Rate Arbitrage
# Current funding rates
coinglass_funding_current(action="rates")
# Find arbitrage opportunities
coinglass_funding_current(action="arbitrage")
Whale Tracking
# Recent whale alerts (Hyperliquid)
coinglass_whale_positions(action="alerts")
# Large BTC positions
coinglass_whale_positions(action="positions", symbol="BTC")
Market Sentiment
# Fear & Greed Index
coinglass_indicators(action="fear_greed")
# Bitcoin Rainbow Chart
coinglass_indicators(action="rainbow")
Tool Discovery
# Search for liquidation-related tools
coinglass_search(query="liquidation")
# Check available features for your plan
coinglass_config(action="plan_features")
Architecture
coinglass-mcp/
├── src/coinglass_mcp/
│ ├── server.py # FastMCP server + 22 tools
│ ├── client.py # HTTP client with retry logic
│ └── config.py # Plan tiers, intervals, features
├── tests/
│ ├── test_client.py
│ └── test_tools.py
└── pyproject.toml
Design Principles:
- 3-file architecture — Optimized for AI agent comprehension
- Domain facade pattern — 22 tools → 80+ endpoints
- Literal-typed actions — Helps LLMs select correct operations
- Lifespan pattern — Shared httpx.AsyncClient for efficiency
Development
Setup
git clone https://github.com/forgequant/coinglass-mcp.git
cd coinglass-mcp
uv venv && source .venv/bin/activate
uv pip install -e ".[dev]"
Testing
pytest -v
======================== 45 passed in 0.69s ========================
Run Locally
export COINGLASS_API_KEY="your-key"
python -m coinglass_mcp.server
FastMCP Cloud Deployment
Entry point: coinglass_mcp.server:mcp
Environment variables:
COINGLASS_API_KEY— Your CoinGlass API keyCOINGLASS_PLAN— Subscription tier (default:standard)
API Reference
Full CoinGlass API documentation: open-api.coinglass.com
License
MIT
Links
- CoinGlass — Cryptocurrency derivatives analytics
- FastMCP — Fast, Pythonic MCP server framework
- MCP Protocol — Model Context Protocol specification
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。