Judging an agent by "how good the final reply sounds" is rarely enough.
What teams actually worry about happens at the tool-call layer. After changing a prompt, swapping a model, or tweaking a tool, the replies may look better while the agent quietly starts to:
hit "approve" on a ticket it hasn't fully read;
call a write endpoint it should never touch;
run "check inventory, then place order" in the wrong order;
trigger an action twice β or, when it should act, do nothing at all.
None of that shows up by reading the streaming reply. Baize evaluates agents on the tool-call timeline instead: not what the model said, but what it actually did β and by default the scoring is deterministic code rules, not an LLM judge.
1.β¦