Start every new agent at under two minutes with zero…
Design minimal viable agents to execute in under two minutes with zero preparation before attempting multi-step sequences, because automaticity requires low activation energy and activation energy must be minimized before sophistication is added.
When This Fires
- When designing any new behavioral agent from scratch
- When rebuilding a failed agent after simplification (When an agent fires below 80% after 30 days, simplify…)
- When someone describes an ambitious new habit — check activation energy first
- When an agent requires preparation steps that themselves have activation energy barriers
Common Failure Mode
Designing the aspirational version first: "My new exercise routine is 45 minutes of lifting, 3x per week." The activation energy (changing clothes, commuting to gym, warming up) is so high that the agent fails within two weeks. The minimal viable version — "Put on gym shoes" or "Do 5 pushups at home" — would have established the trigger-response pathway that the 45-minute version needs but can't build on its own.
The Protocol
(1) When designing a new agent, ask: "What is the absolute minimum version of this behavior that takes under two minutes and requires zero preparation?" (2) That is your starting agent. Not a "stepping stone" or "training wheel" — this IS the agent. (3) Install it: set the trigger (Agent triggers must be observable or measurable — vague…), define the condition, execute the minimal action. (4) Track for two weeks: does it fire at 80%+? If yes → the automaticity foundation is building. (5) After consistent firing, expand by small increments: two minutes → five minutes → ten minutes. Each step must maintain 80%+ reliability before the next expansion. (6) If the minimal version doesn't fire → the issue is trigger or condition design, not action complexity. Diagnose before redesigning (Diagnose failing behavioral agents by component — trigger…).