请问HN：在构建代理时，您会跟踪哪些指标？
我关注的主要指标是推理成本，这通常是判断问题的一个很好的指标。<p>然而，几天前我发现由于一个bug，我们加载了大约300个工具，这影响了成本和准确性。这种情况持续了一段时间，我没有通过查看成本意识到这一点，因为在同一时期我做的其他更改掩盖了这个问题。<p>现在我开始跟踪这个指标和其他一些指标，但这让我想知道还有什么我可能遗漏的。因此，你们在跟踪什么？有什么建议吗？很想听听。
查看原文
The main metric I look at is inference cost, which is a pretty good indicator when something's off.<p>However, a couple of days a go I discovered we were loading ~300 tools due to a bug, affecting both costs and accuracy. It was going on for some time and I didn't realize it by looking at the costs, cause it got hidden by other changes I made in the same period.<p>Now I started tracking that and some other metrics, but it made me wonder what else I might be missing. So: what are you tracking? Any tips? Curious to hear