LLMyth

How LLM Agents Became What They Look Like in 2026?

Tools, MCP, Bash, Filesystem, OS, and Agent Skills

有机的 LLM 人机交互: 不只是对话框

Why Does My LLM Structured Output Perform Poorly?

There is no silver bullet for LLM structured output.

Agents Need “State” Instead of “Memory”

如果我们想要一个解决问题的 “工业机器人” agent，这种 memory 既不符合我们欣赏的软件哲学，在技术上也值得质疑。“工业机器人” agent 需要设计的是 state。

Only Two Methods Needed for Agents -- LLM as a "magic function" & LLM do "function calling"

Trying to build AGI with only two methods!