Repository navigation
[Feature] Long-term metabolic profile and portable reports #48
Description
Outcome
Turn qualified multi-week and multi-sensor data into an explainable long-term glucose/context profile and portable report without inventing an opaque metabolism score or “bio-age.”
Dependencies
- Deterministic evidence and qualification: [Feature] Deterministic metabolic observations and evidence model #37
- Unified timeline and drill-down: [Feature] Unified body timeline and compact Today experience #44
- Complete export/data lifecycle: [Feature] Complete health-data export, deletion, retention, and recovery #40
- Context quality improves as [Feature] Fast journal and meal/activity logging #41, [Feature] Apple Health context import #43, and [Feature] Android Health Connect context import #42 land.
Scope
- Total and recent usable coverage with provenance and gap warnings.
- Rolling glucose average, variability, range exposure, meal response, overnight profile, and movement/workout context.
- Recent versus prior qualified-period comparisons.
- Context-aware weekly recap.
- User-selected date range and included data types.
- Readable PDF/report plus machine-readable evidence attachment.
- Plain-language metric definitions and source attribution.
- Preview/redaction controls before sharing.
Parallelization
Blocked for production integration. Report layout and export-format prototypes may use synthetic evidence fixtures after #37 stabilizes. Do not reimplement analytics in the report layer.
Acceptance criteria
- Every comparison independently qualifies both periods and exposes gaps.
- Reports show source, window, evidence counts, and metric definitions.
- Missing context never becomes a zero or an implied conclusion.
- No opaque health score, bio-age, causality, diagnosis, treatment, dosing, or emergency guidance.
- Identifier minimization and redaction preview work before file creation.
- Generated files are cleaned up and excluded from backup as required.
- PDF/readable output passes accessibility and small/large dataset rendering checks.
- Machine-readable attachment round-trips to the canonical evidence model.
- R2 privacy and independent review are complete.
Activity
shroominic commentedon Sep 1, 2026
Superseded by #101, recreated by the repository owner under
shroominic.