Developers building AI agents, where do you actually want verification to happen?
retained text
I’m building an open-source AI verification tool, and I’ve been thinking about how developers prefer to investigate failures after running tests. Do you prefer having verification results directly inside your IDE, accessing them through an API, or using a separate dashboard where you can inspect evidence, compare test runs, and investigate failures? I’m especially curious about what happens after a test fails. Do you want to investigate it in the same environment where you’re writing code, or would a dedicated verification workspace be useful? Would love to hear how others are handling this today.
document fetched 2026-11-08 00:18 · canonical https://reddit.com/r/AI_Agents/comments/1xq2w9n/developers_building_ai_agents_where_do_you