Decision Framework
The framework converts heterogeneous evidence into one current priority without pretending that every signal is independent, equally reliable or equally useful.
Five Evidence Layers
- A
Stable profile
Age context, goals, dietary pattern, work type, training experience, physical limitations and recurring responsibilities.
- B
Semi-stable baseline
Usual sleep, normal energy, average training frequency, typical working pattern, meal structure and normal social contact.
- C
Dynamic state
Recent sleep, soreness, workload, intake, stress, focus, planned training and immediate priorities.
- D
Behavioural or passive evidence
Action completion, training records, movement data, calendar context, wearables and app behaviour, where available and permitted.
- E
Outcome and correction
Whether the action was completed, whether it helped, whether the classification felt accurate and what the system missed.
Collection rule. Onboarding should cover stable profile and selected baseline information only. Dynamic state, behaviour and outcome evidence must be refreshed after onboarding. Static answers must not be used to infer rapidly changing debts indefinitely.
Causal suppression
An upstream explanation reduces the weight of downstream symptoms. Poor sleep must not independently maximise Sleep, Recovery, Energy and Focus. A missed action must not default to Execution when Decision, Planning or Setup better explains the failure.
Output contract
One primary debt. Optional secondary contributor. Evidence. One proportionate action. One verification method. Confidence and missing data.
Uncertainty rule
Near ties favour the causally upstream debt or trigger one clarifying question. Low confidence triggers clarification or a no-reliable-priority state, not a guess.
Selection Dimensions
- Evidence qualityDirectness and reliability.
- Data recencyHow current the signal is.
- Baseline deviationDifference from the user's normal.
- Severity and trendMagnitude and direction of change.
- Goal relevanceImportance to the present objective.
- Causal priorityAbility to explain other signals.
- UrgencyCost of delaying action.
- ActionabilityFeasibility under current constraints.
- Expected benefitLikely value of intervention.
- ConfidenceStrength of the total case.
The 15 Controlled Debts
Sleep Debt
Insufficient, irregular or disrupted sleep opportunity and continuity.
Recovery Debt
Physical or cognitive load exceeding current recovery capacity.
Downshift Debt
Difficulty reducing mental or physiological arousal when recovery is required.
Energy Debt
Persistently low usable energy not better explained by a supported upstream cause.
Protein Debt
Insufficient protein quantity, distribution or consistency for the user's goal.
Hydration Debt
Insufficient or poorly distributed fluid intake relative to conditions and activity.
Fuel Debt
Insufficient total energy or carbohydrate availability for current demands.
Progressive Overload Debt
Training is consistent but stimulus or measurable progression is absent.
Movement Debt
Non-training movement is insufficient or sedentary exposure is excessive.
Interaction Debt
Meaningful contact is below the user's own need, regardless of interaction volume.
Focus Debt
Attention cannot be sustained after the task is clear and work has begun.
Decision Debt
An unresolved choice or priority prevents commitment to a direction.
Planning Debt
The objective is chosen but the sequence, timing or next action is undefined.
Setup Debt
A plan exists but the environment, materials, cues or access are not ready.
Execution Debt
A clear, feasible and prepared action is repeatedly not initiated or completed.