sandboxing agent actions

[cacheline]

I don't think that's the whole story. I get why that seems right, but on M3 Max the bottleneck felt elsewhere.

I'd start with the simpler stack. Turns out 'just one more test' was a lie.

[fastquill74]

Interesting, I hadn't tried that. I ended up in about the same place using LM Studio. Worth rechecking your defaults though, because one bad assumption can make the rest of the thread look confusing.

If you post your exact settings, people here can probably narrow it down pretty fast.

[circuitworks90]

Good point. I ended up in about the same place using SGLang. I got better results once I stopped chasing one metric and just looked at whether the workflow felt usable.