🎮 GRID.gg Esports MCP Server
A "Brain Extension" for AI Agents to directly access professional Esports data.
This Model Context Protocol (MCP) server bridges the gap between Large Language Models (like Claude, Gemini, or JetBrains AI) and the GRID.gg Esports API. It allows AI assistants to autonomously search for teams, find match replay files, and understand the complex data schemas of professional League of Legends and Valorant matches.
Built for the Cloud9 x JetBrains "Sky's the Limit" Hackathon.
✨ Features
This server exposes 4 powerful tools to your AI Agent:
explore_graphql_schema: The "Librarian." It introspects the GRID Central Data API to learn available fields (Teams, Players, Tournaments) on the fly, ensuring zero hallucinations when writing queries.execute_central_query: The "Scout." Executes live GraphQL queries to find Team IDs, Match IDs (Series), and Tournament details.list_match_files: The "Analyst." Retrieval tool that finds specific JSONL replay files (series-events) for a given match.get_event_definition: The "Dictionary." Provides the hard-coded schema for complex in-game events (likeward-placedorplayer-moved) that aren't available via GraphQL introspection.
🚀 Getting Started
Prerequisites
- Node.js (v18 or higher)
- A GRID.gg API Key (from the Hackathon portal)
Installation
- Clone the repository:
git clone [https://github.com/resilientbeast/grid-esports-mcp.git](https://github.com/resilientbeast/grid-esports-mcp.git)
cd grid-esports-mcp
```
2. **Install dependencies:**
```bash
npm install
```
3. **Build the server:**
```bash
npm run build
```
## ⚙️ Configuration
To use this with an MCP Client (like **Claude Desktop** or **JetBrains AI**), add the following configuration to your config file (e.g., `claude_desktop_config.json`):
```json
{
"mcpServers": {
"grid-gg": {
"command": "node",
"args": [
"/absolute/path/to/grid-esports-mcp/dist/index.js"
],
"env": {
"GRID_API_KEY": "YOUR_ACTUAL_API_KEY_HERE"
}
}
}
}