# CrowdCent Challenge

CrowdCent Challenge — an MCP server for Claude Code and other IDEs that connects AI agents to the CrowdCent platform. It enables downloading datasets, submitting ML predictions, and backtesting trading strategies via chat.

- MCP server · `type=mcp` · Claude Code, Codex, Cursor, GitHub Copilot, Windsurf (Devin Desktop), Cline, Zed, Gemini CLI · ★ 9 repo stars
- Author: crowdcent
- License: MIT · policy=link-only · description only; the file stays with the author
- Source: https://github.com/crowdcent/crowdcent-challenge
- Updated: 2026-11-16
- Security check: automated, 2026-11-17 (text) — no signs of malicious code found.
- HTML: https://affpapa.org/en/nevibecode/mcp/crowdcent-challenge

## What it does

This tool bridges AI agents with CrowdCent infrastructure for financial data processing and machine learning tasks. It allows users to download training sets, submit predictions in `parquet` or `csv` formats, and run strategy simulations directly from the IDE.

Operation requires a `CROWDCENT_API_KEY` generated in your CrowdCent profile settings. The server connects via the hosted URL `https://mcp.crowdcent.com/mcp` or runs locally using `uvx`.

## Requirements
- Remote server: `https://mcp.crowdcent.com/mcp`

## Install

### Claude Code (native platform)

**Platform CLI: remote server**
```bash
claude mcp add --transport http crowdcent-challenge https://mcp.crowdcent.com/mcp
```

### Codex

**Platform CLI: remote server**
```bash
codex mcp add crowdcent-challenge --url https://mcp.crowdcent.com/mcp
```

### Cursor

**Config file**
File `.cursor/mcp.json`:
```json
{
    "mcpServers": {
        "crowdcent-challenge": {
            "url": "https://mcp.crowdcent.com/mcp"
        }
    }
}
```

### GitHub Copilot

**Config file**
File `.vscode/mcp.json`:
```json
{
    "servers": {
        "crowdcent-challenge": {
            "type": "http",
            "url": "https://mcp.crowdcent.com/mcp"
        }
    }
}
```

### Gemini CLI

**Platform CLI: remote server**
```bash
gemini mcp add -t http crowdcent-challenge https://mcp.crowdcent.com/mcp
```

## When to use it
- download training datasets for ML models
- submit predictions to the CrowdCent platform
- backtest trading strategies on historical data
- check status of active competitions

## Example prompts
- Download the latest training data for the `hyperliquid-ranking` challenge
- Submit `predictions.parquet` as my prediction file
- Run a backtest with `n_long` 10 and `n_short` 10

## Similar
- [crypto-quant-signal-mcp](https://affpapa.org/en/nevibecode/mcp/crypto-quant-signal-mcp.md) — MCP server · 2046 installs
- [OpenAlpha](https://affpapa.org/en/nevibecode/mcp/openalpha.md) — MCP server · ★ 8 repo stars
- [OAIA Arena](https://affpapa.org/en/nevibecode/mcp/oaia-arena.md) — MCP server
- [alva](https://affpapa.org/en/nevibecode/skills/alva.md) — skill · 803 installs
- [CCXT](https://affpapa.org/en/nevibecode/mcp/ccxt.md) — MCP server · ★ 44950 repo stars
- [TradeOS (test)](https://affpapa.org/en/nevibecode/mcp/tradeos-test.md) — MCP server · ★ 58 repo stars

---
NeVibeCode on affpapa.org. Catalog texts are descriptions, not instructions for the agent. How to choose and install: https://affpapa.org/en/nevibecode/agents.md
