For the complete documentation index, see llms.txt. This page is also available as Markdown.

G.A.M.E.: Functional AI Agent Framework

Learn how G.A.M.E. (Generative Autonomous Multimodal Entities) provides developers with an AI agent framework, API, and SDK for agent behavior, planning, memory, and onchain operations.

G.A.M.E. functional AI agent framework

G.A.M.E. functional AI agent framework architecture
G.A.M.E. Functional Agent

Generative Autonomous Multimodal Entities (G.A.M.E.) is the first product designed for developers to access and experiment with Virtuals AI agents through an API and SDK.

G.A.M.E. agent behavior and architecture

The Agent Prompting Interface provides access to agentic behavior features. The Perception Subsystem synthesizes messages and sends them to the Strategic Planning Engine. This engine works with the Dialogue Processing Module and Onchain Wallet Operator to generate responses.

The Long-Term Memory Processor extracts relevant experiences, reflections, dynamic personality, world context, and working memory. This information supports AI agent decision-making.

Build functional AI agents with the G.A.M.E. API and SDK

Feedback loops help AI agents refine general knowledge for future planning. Agents evaluate the outcomes of actions and conversations.

G.A.M.E. is a lightweight framework for adding plug-and-play AI agents to projects.

Last updated