Hey everyone. I was trying to mass test my AI agent that I created for a Real Estate company to verify leads. So I've created a test suite for it based on one of the conversation logic in a call. And I've encountered a super weird issue. Which is that the script agent would do weird things from speaking to itself as if a conversation is taking place, or spam User:... (check image for better understaning) I've tried to change the LLMs used, but that didn't work. I tried rewriting the script & rubric, but it also didn't work. Is this because the test suite is being deprecated or something? Thanks. I'd appreciate any help. https://cdn.discordapp.com/attachments/1211484481848873020/1514696069693571307/image.png?ex=6a2c4e24&is=6a2afca4&hm=9865548c8fbebbf3cb2200576cdcc209e084c9c2c525465a9a449bf32de51bee& https://cdn.discordapp.com/attachments/1211484481848873020/1514696070121128078/image.png?ex=6a2c4e24&is=6a2afca4&hm=02a2e295eac21072d3552bb51fe48b9c8c96c85947baf8c686b817e3910c97cd& https://cdn.discordapp.com/attachments/1211484481848873020/1514696070624706730/image.png?ex=6a2c4e24&is=6a2afca4&hm=c76396033b6b536868cd97ac7db747688152c8a34fff7fdf5a532f126630a510& https://cdn.discordapp.com/attachments/1211484481848873020/1514696071077560351/image.png?ex=6a2c4e25&is=6a2afca5&hm=872534ec1ff735a30800870d629c92d0eaf3e97c109011b589ee55ba72830dcb&