It seems that over the past few months, they have implemented a quite special development: AI is no longer just a "research assistant" that helps you look up information, but has directly entered trading decision-making and completed the execution loop. In January, during the @Aster_DEX trading competition, they conducted real-world testing by feeding the AI macro, on-chain, and sentiment data—just to see if this information could actually turn into valuable trading signals.

The result showed that this AI is not limited to using a fixed strategy. Slightly adjusting the structure of the Agent completely changes its behavior, making it more like a "strategy generator." The real difference isn’t in competing with humans on speed (since latency is unavoidable), but in thoroughly understanding fragmented data and capturing some macro trends that others find hard to see.
Now they are gradually releasing more features, including macro perception analysis, multi-source data aggregation, and AI-generated real-time trading signals. They also plan to publish the execution addresses on-chain, making all strategy operations transparent and auditable, then use the entire closed-loop feedback to further optimize the SkyAI ecosystem.
What I see here is that this is no longer just simple "AI-assisted trading," but AI actually stepping in to become a trader.
If you're interested in this area, you can first check out the thread where they clearly explain their ideas—it's a good opportunity to understand the logic while exploring potential.
@SKYAIpro I believe in you, and hope this MCP-driven full-ecosystem AI can create a new world.