In the versions of AI toolkit (after version 12.x) custom evaluators shift from needing version 1.0.1 of azure-ai-evaluation to needing version 1.6.0. if you try to run a custom evaluator with any ...

when the Copilot´s AI tries to run a command, there is a good amount of times in which it doesnt do it properly (uses bad syntax to run a command), and/or doesnt move on when a command executes on the ...