Browse the library

An alphabetical directory of published knowledge, collections and longreads. Every page works without JavaScript.

1154 entries · Page 1 of 12

• Ablate a skill before you trust itprotocol
• Act out simple action phrases to remember themprotocol
• Add light accountability to extensive readingprotocol
• Add one adversarial case the coding agent did not proposeprotocol
• Add planner-builder-evaluator roles only after the simple loop hits a ceilingheuristic
• Add professional support to a quit-smoking attemptprotocol
• Add target constraints after the core decision is stableprotocol
• Aggregate repeated signals before turning them into an identityheuristic
• Agree how the negotiation will run before arguing the termstemplate
• Agree what counts as a fair comparison before choosing the benchmarkchecklist
• Agree what reliability loss will pause ordinary changestemplate
• AI Assistants and Vedokrok: Do You Need a Public Method?longread
• Allocate preparation by weighted gaps, not by comfortheuristic
• Alternate private generation with shared idea reviewprotocol
• Anchoring effectconcept
• Announce the result without stealing the user’s placeprotocol
• Approve the exact action that will actually runchecklist
• Archive inactive knowledge out of the default viewprinciple
• Ask 'and then what?' before taking the first-order winheuristic
• Ask a coding agent for one coherent increment at a timeprinciple
• Ask a few target questions before you studyprotocol
• Ask AI to extend the reasoning, not only answer itpattern
• Ask an open question before assuming the other side's constraintprotocol
• Ask before you mind-readprotocol
• Ask for a cue before the full correctionprotocol
• Ask for a route, not a rescueprotocol
• Ask for advice on the next attemptprotocol
• Ask for bounded help while making no a real optionprotocol
• Ask for disconfirming information firstquestion
• Ask for enough help to keep owning the next stepcollection
• Ask for one next behavior, not a personality renovationprinciple
• Ask for one piece of advice for the next attemptprotocol
• Ask for one stretch responsibility, not a vague bigger roleprotocol
• Ask for the artifact that could outvote memorychecklist
• Ask for the future condition both sides could live withquestion
• Ask for the last real example before asking for the usual processheuristic
• Ask for the smallest approval that unlocks learningprotocol
• Ask how the system will respond to the interventionquestion
• Ask questions that produce usable evidence instead of polished generalitiescollection
• Ask quieter experts before the loudest consensusprotocol
• Ask the learner to generate options before advisingprotocol
• Ask the next question from the last thing they saidheuristic
• Ask what a nonsignificant study could have missedquestion
• Ask what each stakeholder can block or enablequestion
• Ask what helped, what hindered and what changes nextprotocol
• Ask what makes the better behavior hardquestion
• Ask what perfect information could changequestion
• Ask what stopped the near miss from becoming harmquestion
• Ask what the percentage is a percentage ofquestion
• Ask what the position is trying to protectquestion
• Ask what was randomizedconcept
• Ask what would make the claim falsequestion
• Ask whether you are changing a parameter or a rulequestion
• Assume the measurement can change the behaviorprinciple
• Attach confidence to the evidence, not to your toneprotocol
• Attach evidence to every skill you want others to believeprotocol
• Attach the focus block to an observable cueprotocol
• Attach the source next to the claim it supportsprinciple
• Attempt the learning task before opening AIprotocol
• Audit an MCP server before giving it repository contextprotocol
• Audit family technology by interrupted moments, not only total screen timeheuristic
• Audit the reference class before trusting its base ratequestion
• Audit the version before merging an AI-added dependencyprotocol
• Automate the part you can change without waiting for everyone elseheuristic
• Availability heuristicconcept
• Average independent estimates before debating one numberprotocol
• Batch by setup state, not only by task labelheuristic
• Batch low-risk maintenance at natural boundariesprotocol
• Batch low-urgency notifications into predictable windowsprotocol
• Batch repeated work without turning batching into a rulecollection
• Before You Refresh Readwise Readerlongread
• Benchmark human plus AI against the stronger solo modetemplate
• Bind the verification to the exact artifact reviewedprotocol
• Bisect a large failure domainheuristic
• Bisect the version history instead of inspecting every changeprotocol
• Block briefly when the pattern is still formingprinciple
• Block secrets before they become repository historyprotocol
• Book summaries or a method for the task?longread
• Books, summaries, highlights—or methods?longread
• Bound every autonomous agent loop before it startsprotocol
• Bound how much damage one agent run can dochecklist
• Break a dependency before optimizing inside itheuristic
• Break the batch when delay costs more than the switchheuristic
• Brief the team before consequential work startsprotocol
• Bring a criterion, not only a preferencechecklist
• Bring objective traces into the debriefchecklist
• Bring one recommendation and its trade-offprotocol
• Budget two to five months for habit formation when the behavior matterstemplate
• Buffer variability where it is cheaper to absorbheuristic
• Build a developmental bench, not one perfect mentorcollection
• Build a developmental bench, not one perfect mentorprotocol
• Build a morphological box from the design dimensionsprotocol
• Build a preparation rhythm that survives cognitive loadcollection
• Build a preparatory habit around the behavior that keeps failing to starttemplate
• Build a reusable research wiki when the topic keeps coming backprotocol
• Build a task bank for variation, not repetition countprotocol
• Build a terminology map before optimizing the queryprotocol
• Build an evidence-by-hypothesis matrixprotocol
• Build an outside-view reference classprotocol
• Build an overall rate from totalsprotocol

Inspect the public connection graph. Connections explain how pages and sources relate; they are not quality scores.