Getting Started

Installation Guide

MOTO Math Variant V1 is now available as open source software. The system supports Windows, macOS, and Linux with both one-click and manual installation options.

Quick Install (Windows):

1. Install Python 3.8+ and Node.js 16+
2. Download and install LM Studio (optional – system also works with OpenRouter)
3. Clone or download MOTO from GitHub
4. Double-click launch.bat – the launcher handles everything automatically

The one-click launcher automatically checks prerequisites, installs dependencies, starts the backend and frontend servers, and opens the UI in your browser.

Manual Installation: For other platforms or custom setups, follow the detailed platform-specific instructions in INSTALL.md included with the repository. The manual installation guide covers Windows, macOS, and Linux with step-by-step instructions.

Core Documentation Files

MOTO includes comprehensive documentation in the GitHub repository:

README.md – Project Overview

The main README provides a complete overview of the MOTO system, including feature highlights, quick start instructions, and system architecture. Start here for a high-level understanding of what MOTO does and how it works.

INSTALL.md – Detailed Installation

Platform-specific installation instructions for Windows, macOS, and Linux. Includes troubleshooting for common installation issues, verification steps, and alternative installation methods.

QUICKSTART.md – Usage Guide

Step-by-step guide to running your first MOTO research session. Covers the one-click launcher, manual setup, configuration, and monitoring your first aggregation and compilation workflows.

START_HERE.txt – Quick Reference

A quick reference guide with prompting tips, system URLs, troubleshooting, and essential information for getting started quickly. Perfect for experienced users who need a refresher.

CONTRIBUTING.md – Developer Guide

Guidelines for contributing to the MOTO project, including code style, testing procedures, pull request process, and development environment setup.

Advanced Topics

Design Specifications

The .cursor/rules/ folder contains complete system design specifications used by AI development assistants. These documents provide deep technical insight into:

• Multi-agent aggregation workflow architecture
• Paper compilation and validation system
• Autonomous research mode logic and decision-making
• RAG pipeline design and optimization
• File structure and module purposes

If you use Cursor IDE, these specifications enable AI-assisted development and customization of the MOTO system.

Configuration Options

Model Selection: Each role (submitter, validator, compiler agents) can use different models with independent context windows and output token limits.

Context Windows: User-configurable per role (default: 131,072 tokens). Larger windows support more complex research with extensive context.

Submitter Count: Configure 1-10 parallel submitters in Aggregator mode (default: 3). More submitters enable broader exploration of the solution space.

OpenRouter and LM Studio capability: Use premium OpenRouter models from the largest AI providers. MOTO can work with almost any type of LLM, both diffusion AND transformer models working together on one output!

Troubleshooting

Common issues and solutions:

• “Python not recognized”: Reinstall Python and check “Add Python to PATH” during installation
• “Failed to connect to LM Studio”: Ensure LM Studio is running with server started on port 1234, and load the required nomic-ai/nomic-embed-text-v1.5 embedding model
• “Port already in use”: Close other applications using ports 8000 or 5173, or restart your computer
• High rejection rate: Check that models are generating valid JSON, review validator reasoning in logs, ensure prompts are clear and specific
• System running slow: Use faster/smaller models, reduce context window size, or close resource-intensive applications

See INSTALL.md and QUICKSTART.md for comprehensive troubleshooting guides.

GitHub Repository

MOTO is open source software released under the MIT License. The complete source code, documentation, and issue tracker are available on GitHub.

Repository includes:

• Full source code (backend Python + frontend React)
• Complete documentation files
• Design specifications for AI-assisted development
• One-click launcher for easy setup
• Example configurations and workflows
• Issue tracker and discussions

Check the Download MOTO page for the GitHub repository link and installation instructions.