Dice Rolling

About

Addressing the inherent limitation of Large Language Models (LLMs) to perform truly random dice rolls, this Model Context Protocol (MCP) server acts as a crucial bridge, supplying AI assistants with genuine random number generation. While LLMs merely simulate rolls through text prediction, this TypeScript-based server parses standard dice notation, generates cryptographically secure random numbers, applies complex gaming mechanics like advantage, exploding dice, and rerolls, and returns authentic results. This ensures gaming integrity for tabletop RPGs, accuracy for statistical simulations, and reliable random number generation for any AI application.

Key Features

• Offers dedicated MCP tools for dice rolling (`dice_roll`) and notation validation (`dice_validate`)
• 1 GitHub stars
• Supports comprehensive standard dice notation (e.g., 3d6+2, 1d%)
• Implements advanced dice mechanics (Advantage/Disadvantage, Exploding dice, Rerolls, Success counting)
• Designed with a modular and extensible architecture for easy addition of new mechanics
• Provides cryptographically secure random number generation using Node.js `crypto.randomInt()`

Use Cases

• Enabling accurate statistical simulations requiring genuine random number generation
• Integrating true dice rolling into AI assistants for tabletop gaming and RPGs
• Enhancing AI applications that need reliable and verifiable random outcomes