These entries currently meet Brali's reviewed/practical retrieval bar. “Trusted” is a publication state, not a promise that every protocol has equal scientific support.
AI-Assisted PremortemDescribe the plan without sensitive information, ask an AI for plausible ways it could fail, group the suggestions, verify the important ones yourself, and choose mitigations only for risks that survive the check.practical
Ask for Feedback on One Specific DecisionState the intended outcome and the current decision. Ask one focused question such as what is unclear, what assumption looks wrong or what should change first. Choose a reviewer who can judge that question, then record what you will accept, test, discuss, defer or ignore.practical
Judge a Commitment by What Comes NextName the decision. Separate unrecoverable past cost from future cost. List the realistic benefit of continuing, the best alternative use of the same resources, and any obligation or switching cost that genuinely exists now. Then choose stop, continue, change, or gather one missing fact.practical
Record What Would Change Your Confidence Before You CommitState the decision or prediction, label your confidence in ordinary language, list the strongest support and the most important unknown, name one observable condition that would make you update, then revisit the record when the outcome or new information becomes available.practical
Run a Missing-Evidence Check Before a DecisionWrite the example or story currently driving the decision. Ask whether it is representative or simply vivid, recent, or personal. Look for one counterexample and, when the decision warrants it, one relevant denominator, base rate, or source. Update the decision only as far as the new information supports.practical
Run One Disconfirming Check Before a Consequential CommitmentWrite the assumption driving the decision, state one observation that would materially weaken it, identify where that information can be checked, inspect it before the commitment becomes difficult to reverse, and then keep, narrow, change or escalate the plan according to what you found.practical
18 library entries
Full archive below. Evidence states differ; use the trusted list above for normal retrieval and recommendation.
AI-Assisted PremortemUse an AI model as a brainstorming partner for possible failure modes, then verify important risks against the real plan before choosing mitigations.Practical · trusted retrieval
Ask for Feedback on One Specific DecisionGive the reviewer context, ask one focused question, make the reply easy and decide what you will accept, test, discuss, defer or ignore.Practical · trusted retrieval
Challenge Yourself to Step Out of Your Comfort ZoneChallenge yourself to step out of your comfort zone. Ask, 'Is staying the same really better?' Be open to new ways of doing things.Pending review · editorial review required
Consider the Context (Fundamental Attribution Error)When judging someone’s actions, ask yourself, 'What situational factors might be influencing them?' Look beyond just their personality.Pending review · editorial review required
Judge a Commitment by What Comes NextWhen past time, money, or effort is dominating a continuation decision, separate what cannot be recovered and compare realistic future costs, benefits, constraints, and alternativePractical · trusted retrieval
Pause Before Accepting the First Number or FrameBefore committing to the first number, option, estimate, or framing you encounter, make it explicit, generate an independent alternative, and check the information that matters toPending review · editorial review required
Record What Would Change Your Confidence Before You CommitState your current confidence in plain language, record the main support and unknowns, name an observable fact that would change the judgment, then compare the expectation with whaPractical · trusted retrieval
Run a Missing-Evidence Check Before a DecisionWhen a vivid or recent example is dominating a decision, separate what is easy to recall from what is representative, then look for one missing denominator, counterexample, or relePractical · trusted retrieval
Run One Disconfirming Check Before a Consequential CommitmentBefore an ordinary consequential commitment, state the assumption you currently favor, name one observable condition that would weaken it, check that condition directly, and updatePractical · trusted retrieval
Think for Yourself (Bandwagon Effect)Before jumping on the bandwagon, ask yourself, 'Do I really believe in this, or am I just following the crowd?' Make decisions based on your own reasoning.Pending review · editorial review required
Validate Connections (Illusory Correlation)When you notice a connection between two things, ask yourself, 'Is there really a link here, or am I seeing something that isn’t there?' Look for solid evidence before drawing concPending review · editorial review required