5. Products and Features

Omniminds is a full-stack platform comprised of three primary products:

I. The Dojo – Teach by Demonstration

The Dojo is the entry point of the Omniminds ecosystem — a dedicated space where users train agents by recording real digital tasks. Through a simple interface, users demonstrate how to complete tasks using web or desktop applications — whether it’s entering data into a CRM, navigating menus, executing game actions, or filing a report. The system captures:

• Cursor movement and click paths

• Keyboard input

• UI context and on-screen layout

• Application state and transitions

These demonstrations are then used to train CUAs capable of autonomously repeating the task. Contributors are rewarded in $OMNIS tokens based on the quality, utility, and adoption of their demonstrations. Apart from incentivized participation, these demonstrations build the knowledge base for future automation.

II. The Arena – Request & Curate Agents

The Arena enables users, teams, or organizations to request trained agents for specific digital tasks. Whether it’s automating invoice uploads, game grinding, or scraping financial reports, users can define their needs and submit them to the Arena.

From there, Omniminds matches the request with:

• Relevant demonstrations from the Dojo

• Developers who can enhance the agent’s logic

• Validators who ensure performance and accuracy

The Arena acts as a demand coordination layer — connecting needs with crowdsourced agent training in a decentralized workflow. It structures agent specification, organizes matchmaking between trainers and users, and ensures quality assurance before deployment.

III. LaunchBay – Deploy, Run, and Monetize Agents

Once an agent is trained, the LaunchBay is the environment where it comes to life. It allows for deployment, execution, and monetization of CUAs based on real-time user triggers or machine-to-machine interactions.

Agents in the LaunchBay can be activated by:

• Direct user requests (e.g., run a daily data task)

• Integration with other AI agents via Multi-Agent Coordination Protocol (MCP)

• Autonomous behavior schedules or event triggers

• Agent-to-Agent (A2A) payments and delegation

LaunchBay also provides analytics and usage insights for deployed agents, allowing contributors to track performance, optimize behavior, and generate revenue through task execution. Overall, it allows for real-time agent deployment, human/machine activation, and monetization for agent creators.

Together, the Dojo, Arena, and LaunchBay form a continuous learning and automation cycle, where:

• The Dojo captures human knowledge through demonstrations

• The Arena transforms those demonstrations into targeted, validated agents

• LaunchBay brings those agents to life, powering automation and digital labor

This triad turns everyday users into trainers, enterprises into beneficiaries, and the crowd into a decentralized AI workforce.