C.W.K.
Stream
Lesson 05 of 05 · published

Human-in-the-Loop Checkpoints

~26 min · human-in-loop, approval

Level 0Observer
0 XP0/40 lessons0/12 achievements
0/100 XP to next level100 XP to go0% complete

Humans are not just emergency brakes

Human-in-the-loop is often framed as safety approval, but it is also a way to inject judgment, taste, and missing context. Good agents know when a human answer is cheaper than another five tool calls.

The checkpoint should be specific. Ask one decision at a time, present consequences, and carry the answer forward in state.

Checkpoint design

Ask before irreversible actions. Ask when requirements conflict. Ask when the user likely has private context. Ask when the next step spends money or changes external state.

Do not ask vague questions like “what should I do?” Ask “Should I delete these 3 generated files, or keep them for review?”

Code

Checkpoint payload·json
{"type":"approval_request","decision":"delete_generated_files","summary":"Delete 3 generated files under /tmp/export.","options":["delete","keep"],"recommended":"delete","risk":"Deletion is irreversible unless files are regenerated."}

External links

Exercise

Design three human checkpoints for a deployment agent. Make each question specific and consequence-aware.
Hint
One should involve external state, one cost, and one irreversible change.

Progress

Progress is local-only — sign in to sync across devices.
Spotted a bug or have feedback on this page?Report an Issue

Comments 0

🔔 Reply notifications (sign in)
Sign inPlease sign in to comment.

No comments yet — be the first.