Random Tables MCP Server

Random tables have long been a staple of tabletop role‑playing, yet the tools to create and manage them remain fragmented. Random Tables MCP unifies this process by offering a single, modular server that can be accessed either as an LLM‑friendly resource or through explicit tool calls. The goal is to give game designers, writers, and developers a reliable, testable backbone for all kinds of random generation—whether they’re scripting encounters, generating loot, or feeding creative prompts into an AI assistant.

At its core, the server follows a hexagonal architecture. This design keeps the random‑generation logic isolated from the interfaces that expose it, allowing developers to add new access methods without touching the core. As a result, the same table definitions can be rolled in an LLM conversation, queried via REST, or executed by a command‑line script. The separation of concerns also means each table rule can be unit‑tested in isolation, ensuring deterministic behavior even as the system evolves.

Key capabilities include:

• Rapid table creation and updates – Tables can be defined, modified, or deleted on the fly through simple tool calls.
• Instant rolling – A single command returns a roll result, respecting range‑based entries and weighted probabilities.
• Table linking – Templates let you reference other tables, enabling nested or conditional rolls without manual bookkeeping.
• Reusable templates – Standalone roll templates can be stored, listed, and evaluated, making it easy to compose complex generation logic that spans multiple tables.
• Dual access model – Users can choose between MCP Resources (for declarative LLM queries) or Tools (for imperative control), switching via environment variables.

In practice, this server shines in scenarios where unpredictability is a feature rather than a bug. Game masters can generate encounters, loot tables, or NPC profiles on demand; writers can pull fresh plot hooks from a curated set of prompts; developers building AI‑powered content generators can embed the same table logic into chatbots or narrative engines. By keeping the underlying data model consistent across all interfaces, Random Tables MCP ensures that a table created once remains usable and reliable no matter how it’s invoked.

For teams that value maintainability, testability, and flexibility, Random Tables MCP offers a single source of truth for all random generation needs—freeing creators to focus on design while the server handles the dice rolls.