Original Title: After going through 50+ projects in Cookie DeFAI Hackathon, here's what I learned
Original Author: Defi0xJeff, Head of Steak Studio
Original Compilation: Ashley, BlockBeats
Editor's Note: The author summarizes the insights and reflections after browsing over 50 projects from the Cookie DeFAI Hackathon, pointing out the potential and market gaps of DeFAI Agents, especially the rise of vertical Agents and the shortcomings in research Agents. The author believes that Cookie, as a data platform, is driving the development of innovative projects and suggests that teams should focus on new use cases rather than replicating existing functionalities. DeFAI is expected to become an important vertical market in the crypto space, potentially competing with Web2 Agents in the future.
The following is the original content (reorganized for better readability):
After browsing over 50 projects from the Cookie DeFAI Hackathon, here are my takeaways (this is more like feedback / my current views on the Agent market / how projects stand out).
Current Situation: DeFAI = An Abstraction Layer for Many Developers
Many teams added NLP interfaces to their products (possibly because they think DeFAI is equivalent to @HeyAnonai, @griffaindotcom, @orbitcryptoai, @askthehive_ai). In most cases, this is not appropriate—especially when you can only do some simple things, like using the Cookie API to find the influence of the top 5 AI Agent tokens, looking for trending top coins, etc. This is just a mini feature that many top abstraction layers already have.
I believe it would be better to directly use the Cookie dashboard to view these analyses rather than adding a new interface—it is not comprehensive enough.
DeFAI ≠ Abstraction Layer
Instead of replicating existing functionalities, teams should focus on leveraging the Cookie API to unlock new possibilities—driving entirely new use cases and verticals rather than drawing inspiration from existing fields.
The Birth of Vertical Agents
I was surprised by the many interesting ideas that emerged from this hackathon—several projects had unique concepts. Although many projects are still in early demonstration/concept stages, they paint an exciting picture of future use cases.
• An Agent that helps preserve your legacy—checks if you are safe, and if you pass away, it will take action to fulfill your wishes.
• An ETF/index fund for investment decisions and comprehensive research reports using Cookie analytics.
• Agent security analysis and Agent security scoring.
• A ChatGPT-like product/developer learning center that helps developers understand everything about Solana.
• A DYOR layer that tracks calls from analysts/KOLs, DYOR, and copy trading.
• A framework that allows Agents to enter into contracts, enabling complex interactions between Agents or between Agents and humans (unsecured loans, employment agreements, alliances/coordinations).
• Personalized + DeFAI Agent—an AI companion that adjusts its behavior/visuals based on market dynamics.
More and more teams are launching Agents in niche areas, not just "trading Agents" or AI-driven dashboards/research Agents. Launching vertical Agents makes it easier to distinguish them from general Agents.
Trading Agents already have leading players. Although this field is still in its infancy, it remains challenging to stand out—especially in the early stages. Focusing on vertical Agents would be better.
Many might think @HeyTracyAI is @virtuals_io's flagship Agent on Solana, which is useless and cannot help you make money. In fact, an Agent built like a real business—solving real problems—will perform better in the long run. The sports market is a huge total addressable market (TAM). Look beyond Web3. (Not promoting Tracy, just presenting a viewpoint on vertical Agents.)
Conclusion: Niche vertical Agents solve real problems and create unique use cases, while general Agents struggle to stand out.
Lack of Suitable Research Agents
Although vertical Agents are opening up unique niche markets, another major gap in this field is the lack of suitable research Agents.
The key word here is "suitable." Currently, there is no research Agent that can replace human information synthesis and reasoning. This applies not only to the projects from the Cookie DeFAI Hackathon but also to the general situation of Web3 AI Agents.
Most AI Agents today merely aggregate data but do not synthesize insights like humans. Analyzing data through traditional dashboards, such as @cookiedotfun, @GoatIndexAI, @Decentralisedco, and using Grok, is still better than letting AI Agents "feed" Web3 AI Agents "insights."
Despite many abstraction layers and teams focusing on enhancing research capabilities, there remains a clear gap. Whoever can break through this first will gain a significant advantage.
Cookie DeFAI Hackathon Projects
Most hackathon projects are still in early development stages, and many have not yet been deployed. Since this is a purely DeFAI hackathon (as you can see, DeFAI is the best-performing category among AI Agents), many high-quality projects and tokens will emerge from this event.
As mentioned in the second part, many projects will provide new use cases beyond what we currently understand about DeFAI applications.
As AI Agents continue to evolve as a field, Agents can fill more gaps—such as B2A (Business to Agent) going beyond B2B and B2C.
The next wave of DeFAI projects will not only enhance existing use cases—they will create entirely new ones.
Cookie as a Data Support and Distribution Channel for Agents
Unlike relying on launch platforms to highlight unique Agent tokens, Cookie empowers Agents and teams by providing a tracking mechanism for on-chain and off-chain AI Agent data support—enabling new and interesting use cases.
At the same time, Cookie's dashboard is already used by over 240,000 MAU, with users deeply engaged in the field. Discovering gems on the Cookie dashboard and at the Cookie hackathon is like finding a new gem on Virtuals.
Cookie has proven itself to be a powerful distribution channel for Agents. The more Agents leverage this, the faster the ecosystem matures.
Conclusion
This hackathon felt similar to the Solana AI hackathon, but arguably better—because it is purely DeFAI.
DeFAI is not just another AI trend—it has the potential to become the most promising vertical field for Agents in the crypto space. This hackathon has proven that.
I lean towards DeFAI, believing it is a use case for Agents that is native to crypto, capable of developing as an independent vertical field and competing with Web2 Agents.
免责声明：本文章仅代表作者个人观点，不代表本平台的立场和观点。本文章仅供信息分享，不构成对任何人的任何投资建议。用户与作者之间的任何争议，与本平台无关。如网页中刊载的文章或图片涉及侵权，请提供相关的权利证明和身份证明发送邮件到support@aicoin.com，本平台相关工作人员将会进行核查。
