numerai-mcp-autoresearch

rnop/numerai-mcp-autoresearch

★ 0 stars Python 🤖 AI/LLM Updated 2d ago

Agentic research harness for Numerai classic tournament with autonomous experimentation loop with XGBoost, Bayesian optimization, time-series cross-validation, and custom MCP server for weekly retraining and submissions.

View on GitHub → Try with Claude — $10 free →

Quick Install

Copy the config for your editor. Some servers may need additional setup — check the README.

Add to claude_desktop_config.json:

{
  "mcpServers": {
    "numerai-mcp-autorese": {
      "command": "uvx",
      "args": [
        "numerai-mcp-autoresearch"
      ]
    }
  }
}

Or install with pip: pip install numerai-mcp-autoresearch

README Excerpt

GitHub Pages: <a href="https://rnop.github.io/numerai-mcp-autoresearch/" target="_blank">View Research Overview</a> This is an agentic autoresearch and deployment harness for Numerai classic tournament that combines: - Orchestrating agents with instructions in `program.md`, `AGENTS.md` (Codex), and `CLAUDE.md` (Claude Code)

Tools (3)

target_ender_20target_ender_60v52_lgbm_ender20

numerai-mcp-autoresearch

Quick Install

README Excerpt

Tools (3)

Related AI/LLM Servers