How are people debugging multi-agent AI workflows in production?

Re:How are people debugging multi-agent AI workflows in prod... by verdverm on Saturday March 14, 2026 @04:09PM (#47380653)

If you can't trace across agents (like services), then you haven't set up OTEL completely

What your hard fail is, that's at a different layer of control, separate from OP questions about just seeing it so you can design those control systems. That's more guards, validators, and the like (more subagents)

I stay more human in the loop because these things are not ready for prime time the way you describe using them. That's burning tokens on average imo.

[ Parent | on HN ]

Re:How are people debugging multi-agent AI workflows in prod... by skhatter on Tuesday March 17, 2026 @09:04PM (#47420458)

That makes sense — sounds like a lot of this is handled at the framework + design level in your setup.

In practice, when something does go wrong in a multi-step workflow, do you typically rely on tracing + manual debugging, or do you have built-in mechanisms for partial replay / recovery?

[ Parent | on HN ]