Sovrun SDK
The Sovrun SDK ties together traditional game development tools with fully on-chain game and Autonomous World generators
Games like Chain Tactics and Creatorverse run on the game development stack powered by:
a MUD Engine (https://mud.dev/)
a game IDE (Unity/Unreal/etc.) integrator
fine-tuned LLMs for autogeneration of Sovrun mod development and LiveOps
game-specific AI agent templates representing Player Entity(ies).
MUD Engine
MUD was designed by the Lattice team in order to provide an open-source framework for powerful fully on-chain applications such as Autonomous Worlds to be deployed on EVMs. Autonomous Worlds subscribe to the ECS pattern (Entity Component System) used in traditional game development pipelines in order to achieve scalable data architectures and relations, supporting millions of entities at once, optimized for multi-threading and parallelization, through the separation of data and functions.
Game IDE Integrator
Sovrun integrates the MUD Engine with traditional game development environments like Unity, Unreal, Godot, or Bevy. It allows developers to leverage familiar tools and workflows while building MUD-compliant games, enabling seamless integration and deployment of interactive game features.
Autogenerating Fine-tuned LLMs
Fine-tuned Large Language Models (LLMs) assist in the autogeneration of MUD-compliant Sovrun mod development. These models simplify the development process for non-developers and developers unfamiliar with blockchain, automatically producing boilerplate templated code based off of specific game designs. The LLMs are already fine-tuned to specific games in the ecosystem for modular expansion, allowing the longest part of a game's SDLC creation time to be condensed or even entirely skipped past (i.e. the development from scratch of a full game).
Player Entity Game-specific AI Agent Templates
AI agent templates are specifically designed to represent Player Entity(ies) under the ECS concept. These templates operate in a MUD environment to enable autonomous functioning of game-specific AIs, empowering them to move about and interact with smart contracts within the scope of Autonomous Worlds, more efficiently than they would in non-standardized environments (example: games outside of Autonomous Worlds would have to train and integrate agents repeatedly from scratch for each new environment).
For more info on participating in Sovrun's Builder program, visit the Sovrun Discord.
Last updated