Build agentic testing systems to validate AI generated code | ODSP912

AI is driving rapid expansion of software, but vibe coding lacks the rigor needed for reliability. Traditional tests can’t keep pace with agent-generated logic. To scale, we must move beyond manual checks to Agentic Testing: using autonomous agents to validate autonomous systems. Explore patterns for creating autonomous test agents, detecting failures across workflows, and continuously verifying behavior. Take away practical approaches to make AI-driven software reliable and production ready. 𝗦𝗲𝘀𝘀𝗶𝗼𝗻 𝗜𝗻𝗳𝗼𝗿𝗺𝗮𝘁𝗶𝗼𝗻: This is one of many sessions from the Microsoft Build 2026 event. View even more sessions on-demand and learn about Microsoft Build at https://build.microsoft.com ODSP912 | English (US) | Agents & apps Pre-recorded | (300) Advanced #MSBuild Chapters: 0:00 - Exploring the challenge of validating autonomous agent-built applications 00:03:41 - Markdown-based TestMD framework integration for CI/CD 00:04:08 - Shareable evidence through video logs and trace runs 00:05:28 - Starting test case execution and opening KNCLI in interactive mode 00:07:21 - Initializing KNCLI session within agent environment 00:08:35 - Demonstration of KNCLI aiding AI agents in building and testing end-to-end workflows 00:09:23 - Order placement and test data generation for checkout 00:10:02 - First workflow completed and preparing for second step 00:10:45 - Workflow successfully completed and test passed