> For the complete documentation index, see [llms.txt](https://omniminds-ai.gitbook.io/whitepaper/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://omniminds-ai.gitbook.io/whitepaper/4.-who-are-we/crowdsourced-agent-ecosystem.md).

# Crowdsourced Agent Ecosystem

Omniminds enables an open ecosystem of contributors who can create, improve, and validate agents. Each contributor who trains AI agents is compensated with the $OMNI tokens. An AI generates a score for your demonstration quality, which determines how much you’d be incentivized. Every participant has a role:<br>

* Users train agents by demonstrating tasks
* Developers build composable logic modules and fine-tune behaviors
* Validators test and audit agent performance
* Curators organize, optimize, and manage agent libraries
* Communities govern the evolution of agent behaviors

This crowdsourced structure ensures the scalability, diversity, and adaptability of AI behaviors while allowing everyone to earn rewards for their contributions.