The irreducible epistemic atoms underlying the curriculum. 4,828 atoms across 8 types and 2 molecules
Stop work when quality degradation begins rather than when time allocation expires, and use that stopping point to calibrate future capacity estimates.
Track which environmental and behavioral factors predict high versus low capacity days, then design your schedule to increase high-capacity predictors and decrease low-capacity predictors.
Allocate high-capacity days to your most cognitively demanding and irreversible decisions, medium days to execution with clear next steps, and low days to maintenance and recovery.
Consistent moderate output compounds to exceed intermittent high output over time because consistency preserves context, momentum, and decision quality that volatility destroys.
When designing habit replacements, test candidate routines by executing them when the craving hits and waiting fifteen minutes—if the craving resolves, the replacement delivers the real reward; if it persists, keep testing alternatives.
Test candidate keystone habits bidirectionally: genuine keystones improve other domains when present and degrade them when removed.
Set formation timeline expectations of 30-90 days based on behavior complexity before beginning habit installation to calibrate patience as an engineering requirement rather than relying on willpower.
Measure automaticity subjectively on a numerical scale throughout formation to make the logarithmic progress curve visible and prevent interpreting normal midpoint difficulty as evidence of failure.
Use environmental initiation ease as the primary lever for habit installation, since dopamine prediction signals activate at cue recognition rather than reward delivery, making the first step neurologically privileged.
After a single habit lapse, execute the minimum viable version of the behavior the next day to prevent the lapse from becoming a pattern-changing relapse.
Design observation protocols that separate 'how' questions (eliciting description) from 'why' questions (eliciting justification) to collect accurate data rather than defensive explanations.
Configure notification settings and app placement to eliminate externally imposed cues from attention-optimizing platforms, restricting interrupts to person-initiated communications rather than platform-generated triggers.
Design every habit with a degraded mode that preserves the behavior during low-capacity periods rather than allowing complete gaps that reset automaticity.
Grant equal identity credit to both full and degraded versions of a habit to maintain the self-signal that preserves behavioral continuity across capacity fluctuations.
Evaluate each habit quarterly against identity alignment, time ROI, automaticity level, and context fit to detect degraded or obsolete behaviors before they accumulate as behavioral debt.
Increase friction on unwanted routines just enough to create a conscious decision point where automaticity would otherwise execute, buying time to redirect toward replacement behaviors without triggering suppression-without-replacement patterns.
When desired habits conflict with social group norms, either change the social environment, negotiate explicit norm modifications with the group, or create parallel social structures that reinforce the new behavior.
Use goals to set directional targets but rely on systems of daily practices for progress, because goals create binary pass-fail states while systems generate continuous forward motion.
Map reinforcing connections between habits to identify which behaviors feed energy, motivation, or information to others, transforming isolated habits into an integrated system with feedback loops.
Deploy new habits sequentially rather than simultaneously, because each habit in formation consumes the willpower it will eventually conserve, and parallel deployment exhausts available resources before habits stabilize.
Define routines as closed behavioral scripts with no decision nodes—complete sequences where every physical action, location, duration, and completion criterion is specified in advance.
When initial hypotheses form during debugging or incident response, enforce a mandatory observation period before hypothesis testing to prevent confirmation bias from corrupting data collection.
Use the completion of an anchor behavior (the trailing edge) rather than the behavior itself as the cue, because completions are discrete temporal events that provide sharp triggers for pattern-matching systems.
Prioritize controllable cue types (time, location, preceding action) over uncontrollable ones (emotional state, other people) when designing new habits, because reliability requires engineering cues you can influence.