The irreducible epistemic atoms underlying the curriculum. 4,828 atoms across 8 types and 2 molecules
Before building any agent, explicitly name the schema it operates on by writing what the agent assumes about how the world works, because unexamined schemas produce systematically wrong outputs despite reliable execution.
After any difficult social interaction, complete a 5-minute written debrief comparing what your social agent prescribed versus what you actually did, documenting the gap to update the agent based on real behavioral data.
After each decision agent activation, update the agent's criteria based on whether they produced a good outcome, treating the agent as a living heuristic that improves with each use rather than a permanent law.
For sleep agents, set a hard time-based trigger (like 9:30 PM) rather than a subjective state trigger (like 'when I feel tired'), because fatigued brains systematically misjudge their own tiredness.
Build nutrition agents at the meal preparation layer (Sunday evening meal prep) rather than the consumption layer (what to eat when hungry), because deciding what to eat while hungry and facing an open refrigerator is too late to override depletion.
For stress management agents, use physiological signals (chest tightening, jaw clenching, shallow breathing) as triggers rather than cognitive assessments ('when I feel stressed'), because stress impairs the self-assessment capacity needed to recognize stress.
When discovering that your designed agents conflict with each other, resolve the conflict through documented priority hierarchies rather than case-by-case deliberation, making the resolution rule itself part of your agent system.
Build feedback loops into agent systems through regular review asking whether agents fired, whether they produced intended outcomes, and whether conditions have changed, treating review as essential maintenance not optional improvement.
Design triggers using the camera test: if a video camera could not detect the exact moment the trigger fires, the trigger is too vague to fire reliably and must be replaced with an observable event.
When a trigger depends on detecting an internal state, classify it as high-risk and either replace it with an external observable event or invest in deliberate interoceptive calibration before trusting it operationally.
Place environmental trigger objects at eye level or in the direct path of existing routines rather than in convenient storage locations, maximizing visibility over proximity.
Protect the first five consecutive executions of any new time-based trigger as non-negotiable, treating early repetitions as infrastructure investment that determines whether the trigger becomes automatic or dies silently.
Stack temporal triggers with spatial and state anchors (specific time plus specific location plus specific preceding action) to create redundant activation pathways that survive if any single trigger component fails.
Remove or relocate objects that afford unwanted behaviors before adding objects that afford desired behaviors, because elimination removes temptation entirely while addition only competes for attention.
When an environmental trigger requires daily resetting or maintenance to remain visible, replace it with a static persistent cue that survives without intervention, because maintenance costs compound and cause trigger decay.
When designing emotional trigger agents, ensure the action is executable within the emotional state itself rather than requiring the emotion to already be regulated.
Start behavioral triggers with a more conservative threshold than feels right, aiming for 3-5 activations per day rather than 30, to build trust through relevance before expanding sensitivity.
Log each trigger firing for one week as true positive or false positive, then adjust the threshold only after accumulating empirical data rather than based on single instances.
Add qualifying guard clauses to behavioral triggers when false positives exceed 30% of total activations, inserting context checks that must pass before the main action executes.
Match the number of qualifying conditions to the cost of false positives—use minimal guards for low-cost triggers and multiple defensive checks for high-consequence triggers.
When a trigger consistently fails to fire, increase its signal strength through modality shift, physical obstruction, temporal isolation, implementation intentions, or redundant cueing rather than attempting to 'remember better.'
Position physical trigger cues at the exact decision fork where competing behaviors diverge, not merely in the same room, because triggers must intercept choice at the moment it occurs.
Audit your actual movement patterns through physical space before placing any environmental triggers, because triggers placed on aspirational paths rather than real paths will never fire.
For each digital trigger, encode the complete action protocol in the notification text itself ('Daily Review — 3 wins, 1 lesson, tomorrow's top 3') rather than generic labels, to eliminate decision-making at activation.