One repeatable thinking cycle
The program is designed to build a habit of disciplined inquiry rather than reward arrival at one software-approved answer.
STOP
Separate what happened from what you think it means. Preserve the ability to say, “I do not know yet.”
DEFINE
Remove blame and assumed causes. State the actual condition and the decision requiring attention.
QUESTION
Clarify terms, ask how the claim is known, expose assumptions and identify what would change the conclusion.
VERIFY
Classify facts, reports, inferences, assumptions, unknowns and AI-generated material. Trace important claims to original evidence.
CHALLENGE
Use competing hypotheses, contrary evidence, premortems, Devil’s Advocacy and alternative perspectives.
CHOOSE
Compare realistic options. A strong conclusion may be to act, test, wait or escalate.
CONSEQUENCES
Ask what happens next, what happens because of that, who is affected and how behaviour may change.
ACT
Match action to authority, consequence, reversibility and evidence. Preserve meaningful human responsibility where AI is involved.
REVIEW
Return after the outcome. Which assumptions held? What failed? Was the process sound even if the outcome was bad?