Sleep timing varied more
Available nights were less consistent than the proposed comparison window.
VALIDATE FIRST · DOCUMENTATION-ONLY CONCEPT
A weekly decision,
with the evidence attached.
Health Decision Brief is a proposed iPhone experience for adults who already collect fitness data in Apple Health. It would show what changed, how reliable the available evidence is, and up to three small actions to consider—in under five minutes.
No app exists yet. No signup is sent anywhere. This page tests whether the weekly decision problem is worth solving.
01 / THE PROBLEM
The data is already there.
The decision is still yours to assemble.
Today’s workaround is a mix of Apple Health, wearable dashboards, memory, spreadsheets, and general AI. The user still has to decide whether a week is comparable, which change matters, what to try, and whether it helped.
This is a hypothesis, not a research finding. The independent review found no interviews, observed workflows, retention data, or willingness-to-pay evidence.
many metrics→hidden gaps→another score→what now?
02 / THE PROPOSED LOOP
Quality first.
Then interpretation.
The concept begins with approved, available data—not with a confident answer. Every step can narrow the result or stop it.
- 01
Approve
Choose supported Apple Health categories. Read access only.
- 02
Check
Inspect coverage, freshness, provenance, and source changes.
- 03
Brief
See a small set of explainable findings—or an honest abstention.
- 04
Decide
Accept, adapt, dismiss, or pause each proposal yourself.
- 05
Reflect
Record “helpful,” “not helpful,” or “unsure” without claiming cause.
03 / EXPERIENCE SPECIMEN
A brief, not a dashboard
Inspect the trust model.
9:41
WEEKLY BRIEF · CONCEPTUAL
Two patterns are worth a closer look.
◒ PartialData through Sunday
- Coverage
- 6 of 7 nights
- Compared
- Current week vs 3 eligible weeks
- Source
- 1 Apple Health source class
Illustrative values only. Thresholds have not been selected or validated.
PROPOSAL 01Small effort
Keep one wake-time window for four days.
COVERAGE · BEFORE THE BRIEF
What could be compared
Sleep6 of 7 nightsReady ✓
Activity7 of 7 daysReady ✓
Resting heart rate2 of 7 daysOmitted !
Source changed
Activity appears to come from a different source class this week. A real product would ask you to review it.
ABSTENTION · COMPLETE STATE
Not enough comparable data yet.
Nothing is treated as zero, and no action is invented to fill the page.
What we could say
Sleep data was available for 2 of the 5 nights this conceptual method would need.
What happens next
Review coverage, check Apple Health, or return after more history is available.
The numbers shown are layout examples, not product thresholds.
01
Evidence is adjacent
Coverage, comparison window, provenance, and method stay attached to each material claim.
02
A proposal is not a commitment
Only the user can accept or adapt an action. Original wording remains inspectable.
03
Absence is a real answer
When evidence is sparse, stale, contradictory, or unsafe, the brief gets smaller—or stops.
04 / DIFFERENCE, IF PROVEN
Not another score.
A decision record.
Weekly recaps, correlations, reports, and AI coaching already exist. This concept is only distinct if people value the combination below enough to change their weekly behavior.
Existing category
Dashboards, summaries, readiness scores, open-ended coaching
Proposed wedge
Decision compression, claim-to-source lineage, safe abstention, and action/outcome history
Truth today
Plausible, easy for incumbents to copy, and not yet validated with target users
05 / TRUST BOUNDARIES
The product earns the right to be quiet.
Narrow by design.
If research supports a build, the first product would be accountless and local-first on iPhone. It would use selected Apple Health data types and keep the core loop available without a cloud health-data sync.
Designed to do
Read approved categories, check quality, show evidence-linked nonclinical proposals, preserve decisions, export, and delete app-owned data.
Not designed to do
Diagnose, treat, monitor emergencies, change medication or supplements, write generated measurements, publish publicly, or surveil for employers or insurers.
Privacy posture
No account or cloud health sync in the proposed first release. No ads or data sale. Telemetry off until explicit opt-in and review.
This concept is for fitness and wellness reflection, not medical advice, diagnosis, treatment, or emergency monitoring. If someone needs medical or urgent help, this is the wrong product.
06 / AGENTS & MCP
Separate hypothesis. Later gate.
A toolbelt with three locked drawers.
Agent connectivity is not part of validation or the proposed first release. It could be considered only after the standalone loop passes its gates and a specific health/fitness-service recipient passes Apple, privacy, legal, security, and technical review.
07 / VALIDATION, NOT SIGNUP
THE QUESTION THAT MATTERS
Does this solve a weekly decision you actually have?
The next step is research: observe current workflows, compare this artifact with existing tools, test data coverage, and follow the burden over four weeks. A polished page is not evidence.
- Do people describe the problem without prompting?
- Does the brief improve comprehension or task success?
- Will people use an evidence-linked action repeatedly?
- Does abstention build trust—or just add friction?