{"type":"video","version":"1.0","html":"<iframe src=\"https://www.loom.com/embed/1a6fc484dc93431f801e0e6d47d5a446\" frameborder=\"0\" width=\"1108\" height=\"831\" webkitallowfullscreen mozallowfullscreen allowfullscreen></iframe>","height":831,"width":1108,"provider_name":"Loom","provider_url":"https://www.loom.com","thumbnail_height":831,"thumbnail_width":1108,"thumbnail_url":"https://cdn.loom.com/sessions/thumbnails/1a6fc484dc93431f801e0e6d47d5a446-b130762bdda76ec9.gif","duration":257.451,"title":"Crash Lab QA Platform for AI Agents","description":"This Loom explains Crash Lab, a QA assurance platform for testing AI agents built on Confluence Flowwise and Diffie AI. It addresses concerns about agent safety, data security, and robustness against threats like prompt injection, as well as handling hallucinations and load. The platform creates evaluation test cases tailored to the specific agent, with demo cases for a simple chatbot. It then runs executions, shows live run logs, and lets users export results in specific formats after a successful run."}