They don't want to be knee-deep in a debugger.

他们不想深陷试器的泥潭。

Let me go ahead now and get out of the debugger.

现在让我继续，离开试器。

And we'll see one other feature of the debugger today.

今天我们将试器的另一个特性。

But if you check out the Entity Debugger, we now have three entities.

但如果您查实体试器，我们现在拥有三个实体。

And that's what you see in the Entity Debugger.

这就是您在实体试器中的内容。

And that's why the debugger has jumped to that line.

这就是试器跳这一行的原因。

So let's check it out in the Entity Debugger, save the script.

那么让我们在实体试器中检查一下，保存脚本。

We can only see it in the entity debugger.

我们只能在实体试器中。

I recently was reminiscing about how I used to use the debugger so much more before TypeScript made that.

我最近在回想，在TypeScript让我不太需前，我以前是多么频繁地使用试器。

The point is the debugger led you to this point.

关键是试器把你带了这一点。

How do I fix the logic now thanks to the debugger having led me down this road?

由于试器引导我走上了这条路，我现在如何修复逻辑？

For example, in a development capacity, if I wanted to build an expert debugger MCP server, I might have some tools that allow you to do things like get a dump.

例如， 在开发能力方面，如果我想构建一个专家级试器MCP服务器，我可能会拥有一些工具， 这些工具允许你执行诸如获取转储类的操作。

Probably should use I think John Carmack was on here and talked how great the debugger is different world because when I was at Netflix, there's machines that exist somewhere we're on AWS.

我认为可能应该提的是， John Carmack曾在此处谈及试器的卓越性能， 这确实是一个不同的世界。 因为在我在Netflix工作时，尽管我们依赖于AWS上的某些机器， 但我并未直接登录这些设备上。

And so it can effectively download that prompt and then submit that prompt along with all the tools that I got to the LLM to teach the LLM how to become an expert debugger.

因此，可以有效地下载该提示，然后将该提示连同我拥有的所有工具一起提交给LLM，以教会LLM如何成为一个专家试器。