Lean manufacturing's stop-the-line rule offers a framework for deciding when AI agents should stop and ask

Background_Plate1164 · reddit · 2026-11-10

Drawing on continuous-improvement work at a factory in Poland, the author maps lean manufacturing's stop-the-line rule onto the agent question of when to pause and ask a human.

Key points

Agents that pause too often train humans to click "approve" without reading; agents that pause too late let errors compound silently.
Asking the agent "are you sure?" doesn't help much — models are bad at knowing when they're wrong.
The lean-style fix: explicitly define what "normal" looks like and check against that standard, rather than relying on model self-assessment.

The author invites discussion on fixed checkpoints vs. confidence scores and is networking with agent builders in SF.

Original post →

Lean manufacturing's stop-the-line rule offers a framework for deciding when AI agents should stop and ask

More from coding & agent