Inside one turn: reason, then act
Each turn it writes a private thought, picks ONE tool call, then reads the result.
1Reveal the agent's thought, then pick ONE action. Watch the result flow back.your turn
the task
the test for parseDate is failing
FAIL parseDate.test.ts · 1 failing
loop tape, one turn
→ continue← backR replay
If it reasons each turn, should it plan the whole job first, or feel its way?
6.12 The self-correction loop: test, fix, repeatBuilds on6.1The harness: the loop, made real
Common questions
What is "Inside one turn: reason, then act" about?
Each turn it writes a private thought, picks ONE tool call, then reads the result.
What problem does it solve?
You've seen the loop turn. But what does the agent actually do on a single turn?
What will I be able to do after this lesson?
You can explain the ReAct loop: reason → act → observe each turn, and why agents 'think' before each tool call.
What comes next?
If it reasons each turn, should it plan the whole job first, or feel its way?