Gennadiyev/STS2MCP

Full agentic runs for Slay the Spire 2. A mod that exposes in-game state, and the MCP server for the mod.

What it solves

STS2MCP provides a way for AI agents to play Slay the Spire 2, a complex deck-building game. It transforms the game's internal state and available actions into a format that an AI can understand and execute, removing the need for visual-based GUI agents which are often unreliable for gaming.

How it works

The project consists of a game mod that injects a localhost REST API into the game. This API exposes the current game state, player profiles, and a searchable wiki for cards and relics. To make this easier for AI clients, it includes an optional Model Context Protocol (MCP) server that acts as a bridge between the game's API and AI agents (like Claude Desktop or Claude Code), allowing the agent to call tools like get_game_state to make decisions.

Who it’s for

  • AI Researchers: Those testing the reasoning and decision-making capabilities of LLMs in complex, "out-of-distribution" gaming domains.
  • AI Agent Developers: Developers building agents that can interact with software via APIs rather than pixels.
  • Players: Users who want to experiment with AI co-op or automate the game's grind.

Highlights

  • Full Game Control: Supports singleplayer and multiplayer (co-op) runs, including menu navigation, character selection, and lobby management.
  • Durable Context: Provides access to the active profile's compendium, run history, and a fuzzy-searchable wiki for discovered items.
  • Cross-Platform: The mod is a .NET assembly that works on Windows, Linux, and macOS.
  • MCP Integration: Native support for the Model Context Protocol, making it plug-and-play with compatible AI clients.

Related

  • Project
    korotovsky/slack-mcp-server

    An MCP server that enables AI models to interact with Slack workspaces, allowing them to read, search, and manage messages, users, and channels.

  • Project
    classfang/ssh-mcp-server

    An MCP server that allows AI assistants to securely execute remote SSH commands and transfer files while keeping credentials isolated from the AI model.

  • Project
    IvanMurzak/Unity-MCP

    An AI-powered game development assistant that connects AI agents to Unity via the Model Context Protocol to automate editor workflows and enable in-game AI behavior.

  • Project
    miscusi-peek/cheatengine-mcp-bridge

    A bridge that connects AI agents to Cheat Engine via the Model Context Protocol, enabling natural language interaction with program memory for reverse engineering and modding.

  • Project
    ChiR24/Unreal_mcp

    An MCP server and C++ plugin that enables AI assistants to control Unreal Engine, automating tasks from asset management to cinematic sequencing.