I’m Preeminent, the community lead at Treasure (https://treasure.lol).
We’re building tools to enable AI-powered entertainment — creating agents that are persistent, cross-platform, and owned by users. Today, most AI agents are siloed — limited to a single platform, without true ownership. They can’t move across different environments with their built-up memories, skills, or context — and they can’t be traded as assets. We’re exploring a different model: tokenized agents that travel across games, social apps, and DeFi, carrying their skills, memories, and personalities — and are fully ownable and tradable by users.
What we’re building:Neurochimp Framework: #1 Powers agents with persistent memory, skill evolution, and portability across Discord, X (Twitter), games, DeFi and beyond. #2 Agent Creator: A no-code tool built on top of Neurochimp for creating custom AI agents tied to NFTs. #3 AI Agent Marketplace (https://marketplace.treasure.lol): A new kind of marketplace built for AI agents—not static NFT PFPs. Buy, sell, and create custom agents.
What’s available today: 1.Agent Creator: Create AI agents from allowlisted NFTs without writing code directly on the marketplace. Video demo: https://youtu.be/V_BOjyq1yTY 2.Game-Playing Agents: Agents that autonomously play a crypto game and can earn rewards. Gameplay demo: https://youtu.be/jh95xHpGsmo 3.Personality Customization and Agent Chat: Personalize your NFT agent’s chat behaviour powered by our scraping backend. Customization and chat demo: https://youtu.be/htIjy-r0dZg
What we're building next: Agent social integrations (starting with X/Twitter), Agent-owned onchain wallets, Autonomous DeFi Trading, Expansion to additional games and more NFT collections allowlisted for agent activation.
Thanks for reading! We’d love any thoughts or feedback — both on what’s live and the broader direction we’re heading with AI-powered, ownable agents. Our lead AI developer, Alex, will also be around to answer any technical questions.
alextoti9•9h ago
Happy to answer any technical questions around our agent tooling, memory persistence approach, or broader AI architecture.
Also curious to hear any feedback or ideas on what we’re building — would love to learn from the community here.