> For the complete documentation index, see [llms.txt](https://docs.hyperbot.network/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.hyperbot.network/official-links/hyperbot-mcp-server/quick-setup.md). # Quick Setup ### 1. For OpenClaw Bot Users Simply provide the following Skill URL to your OpenClaw Bot. It will automatically fetch the tool descriptions and install the MCP server: URL: **`https://clawhub.ai/developHyperbotNetwork/hyperbot-quote-mcp`** Manual Setup: Go to `Settings` → `MCP` → `Add Server`: * Name: `hyperbot-trading` * URL: `https://mcp.hyperbot.network/mcp/sse` *** ### 2. For Claude Desktop Users Prerequisite: Node.js 18+ installed. Config Path: \* macOS: `~/Library/Application Support/Claude/claude_desktop_config.json` * Windows: `%APPDATA%\Claude\claude_desktop_config.json` Configuration: JSON ``` { "mcpServers": { "hyperbot-trading": { "command": "npx", "args": ["-y", "mcp-remote", "https://mcp.hyperbot.network/mcp/sse"] } } } ``` *Restart Claude Desktop after saving.* *** ### 3. Custom Integration for AI Agents & Apps If you are developing your own AI Agent or application, you can integrate the Hyperbot MCP directly using the following parameters: | **Parameter** | **Value** | | ---------------- | ------------------------------------------ | | SSE Endpoint | `https://mcp.hyperbot.network/mcp/sse` | | Message Endpoint | `https://mcp.hyperbot.network/mcp/message` | | Protocol | JSON-RPC 2.0 | For clients supporting direct SSE URLs: JSON ``` { "type": "sse", "url": "https://mcp.hyperbot.network/mcp/sse" } ``` For clients requiring `mcp-remote` (e.g., Claude Desktop): JSON ``` { "command": "npx", "args": ["-y", "mcp-remote", "https://mcp.hyperbot.network/mcp/sse"] } ``` > Note on Session Management: First, send a `GET` request to the SSE endpoint to retrieve your `sessionId`. All subsequent message requests must include the `?sessionId={your-session-id}` parameter. *** ### Full Capabilities Overview Hyperbot MCP provides 22 data tools, while Skills guide the AI Agent on when to invoke them. Together, they cover 6 dimensions of on-chain game theory analysis on Hyperliquid. **1. Smart Money Discovery** | **Tool** | **Description** | | -------------------- | ------------------------------------------------------------------------------ | | `fetch_leader_board` | Retrieve Smart Money rankings by profit or win rate (24h / 7d / 30d). | | `find_smart_money` | Filter addresses by 9+ metrics (Win rate, ROI, Balance, etc.) with PnL curves. | **2. Market Intelligence** | **Tool** | **Description** | | ------------------- | ---------------------------------------------------------------- | | `get_tickers` | Get latest prices for all supported assets. | | `get_ticker` | Get the latest price for a specific asset. | | `get_klines` | K-line data with buy/sell volume (Intervals: 1m to 1d). | | `get_market_stats` | Order book Long/Short stats, whale order ratios, and mid-prices. | | `get_l2_order_book` | Access full L2 order book depth. | **3. Whale Monitoring** | **Tool** | **Description** | | ------------------------- | ------------------------------------------------------------------ | | `get_whale_positions` | Live whale positions filtered by direction, PnL, or funding rates. | | `get_whale_events` | Real-time stream of whale open/close events. | | `get_whale_directions` | Aggregate whale Long/Short counts (filterable by asset). | | `get_whale_history_ratio` | Historical Long/Short ratio trends (1h / 1d granularity). | **4. Trader Profiling** | **Tool** | **Description** | | ------------------------- | -------------------------------------------------------------- | | `fetch_trade_history` | Complete historical transaction details for a specific wallet. | | `get_trader_stats` | Core stats: Win rate, PnL, holding duration, and trade count. | | `get_max_drawdown` | Max drawdown analysis for 1 to 90-day intervals. | | `get_best_trades` | Retrieve the highest-profit trades within a specific period. | | `get_performance_by_coin` | Win rate and PnL breakdown categorized by specific assets. | **5. Position History** | **Tool** | **Description** | | ----------------------------------- | ----------------------------------------------------------- | | `get_completed_position_history` | Full historical lifecycle of closed positions. | | `get_current_position_history` | Real-time state changes of active positions. | | `get_completed_position_executions` | Exact buy/sell execution trajectories for closed positions. | | `get_current_position_pnl` | Live PnL curve for current holdings. | **6. Bulk Data Queries** | **Tool** | **Description** | | ------------------------ | ---------------------------------------------------------------- | | `get_traders_accounts` | Batch query account overviews for up to 50 addresses. | | `get_traders_statistics` | Batch query trading stats and PnL curves for up to 50 addresses. | *** ### Data Refresh & Limits | **Data Type** | **Frequency** | | ----------------------------- | ------------------------------ | | Market Data / Whale Positions | Real-time | | Trader Statistics | Every 5 minutes | | Smart Money Leaderboard | Hourly | | Rate Limit | 100 requests / minute (per IP) | *** ### FAQ **What is the difference between Hyperbot MCP and Hyperbot Data API?** While both share the same underlying data, they serve distinct use cases. MCP + Skills is tailored for AI Agent scenarios, requiring zero coding and powered entirely by natural language. The Data API is built for Developers, providing comprehensive interface coverage (including real-time WebSocket streams, liquidation data, TWAP, etc.), making it ideal for high-frequency quantitative systems and full product integrations. **Must Claude Desktop users install Node.js?** Yes. Claude Desktop connects to remote MCP servers via the `mcp-remote` bridge, which requires Node.js 18+ to function. macOS users can install it via `brew install node`, and Windows users can download the installer directly from [nodejs.org](https://nodejs.org). **Do I need to install Skills separately for Cursor?** The MCP connection configuration (`mcp.json`) is the mandatory step that enables Cursor to recognize Hyperbot tools. Skills (`npx skills add`) is an optional but highly recommended supplement—it provides the AI with "contextual intelligence" and detailed usage instructions, helping the AI select and invoke the correct tools with significantly higher accuracy. **What is MCP (Model Context Protocol)?** MCP is an open protocol introduced by Anthropic that allows AI assistants to interact with external data and third-party tools through a standardized interface. Hyperbot MCP leverages this protocol to package complex on-chain game theory data into a "plug-and-play" toolset for modern AI. **Which chains are covered?** Currently, Hyperbot MCP focuses exclusively on the Hyperliquid ecosystem, providing deep on-chain data including Smart Money leaderboards, Whale positions, market analytics, and comprehensive trader profiling. **Are there any usage limits?** Yes. To ensure stability, we enforce a rate limit of 100 requests per minute per IP. Additionally, batch query tools support a maximum of 50 addresses per single request. *** ### Related Links * Official Website: * Hyperbot Skills (GitHub): * Hyperbot MCP (ClawHub): * Hyperbot Data API Documentation: