Kiteai @GoKiteAI Important Reminder: Link your EOA wallet If you logged in using a social account (like Email, Google, Apple, GitHub, etc.), make sure to link an EOA wallet (such as MetaMask, OKX, Coinbase, Bitget, etc.). @KiteAIChinese If you initially logged in with a wallet, linking is optional. You can continue using your current wallet or link a new one as needed. Steps to link: 1️⃣ Open the Kite Ozone testnet https://testnet.gokite.ai/ 2️⃣ After logging in, click the avatar or address in the top right corner 3️⃣ Select "Bind EOA" in the dropdown menu and follow the instructions to complete the linking Please link your wallet ASAP! This is a crucial step to ensure your testnet points, rewards, and identity records are properly associated moving forward. —————————————————————— Also, Kite AI @GoKiteAI shared an interesting study: How multi-agent systems (MAS) learn to "plan." In simple terms, it’s about how AI agents can collaborate and divide tasks more intelligently. They introduced a new method: Agent-Oriented Planning (AOP). It enables the meta-agent (main control agent) to break down tasks, allocate them, and integrate results more effectively. The issue with many current multi-agent systems is: Tasks are broken down too much, poorly distributed, and often redundant. AOP aims to solve this. AOP has three core principles: Solvability: Each sub-task can be completed independently; Completeness: All sub-tasks combined can solve the original problem; Non-Redundancy: No repetition, no excess. Under these three rules, the AOP process works like this: → User asks a question → Meta-Agent breaks down the task → Detector checks for redundancy or gaps → Reward Model predicts feasibility → Agents execute tasks → Continuous feedback and optimization In experiments, the research team used GPT-4o as the meta-agent to coordinate four types of sub-agents: math, search, code, and common sense. Results showed: It was 10% more accurate than single-agent systems, 4% better than regular multi-agent systems, slightly more costly, but much more stable. Future directions are clear: 1️⃣ Enable agent-to-agent collaboration; 2️⃣ Introduce human-in-the-loop participation; 3️⃣ Build trust and verification mechanisms, Especially when agents can spend real money, like booking flights or paying for services. Finally, Kite AI also shared their vision: To build infrastructure that allows agents to truly act independently while being trustworthy. This includes identity systems, payments, governance, and verification. Essentially, laying the foundation for the entire AI Agent economy.