Metrics center
Workload Forecast
Public explanation of the live metric contract. Real values appear only inside the user's account and depend on enough evidence.
Question answered
Does future workload fit real study capacity?
Simple interpretation
Forecast shows planned demand and capacity. It is an operational projection, not completed review.
Public technical formula
Versioned operational projection in the public contract.
What counts
Planned items, due and future reviews, weekly capacity, busy blocks, and confirmed optional movements.
What stays out
Future reviews as if already completed, imported calendar titles/descriptions, history before the contract, and missing day rows as zero.
Window and timezone
Main readings use 7, 30, or 90 day windows when applicable. Civil-day aggregates use the authoritative IANA timezone from the profile; rolling timestamps do not depend on the database session timezone.
Minimum evidence
The metric needs enough sample, coverage, or eligible cards. When the canonical contract returns `insufficient_evidence`, `empty`, `stale`, `error`, or `not_applicable`, the UI must show that state instead of inventing a value.
Low-evidence state
Low evidence means the reading is provisional. It can suggest caution, but it is not statistical precision or an outcome promise.
Synthetic example
Synthetic example: if 45 minutes moved through Safe Ahead, workload leaves the source day and enters today without vanishing from the horizon.
Why it changes
Availability changes, new reviews, released cards, confirmed Safe Ahead, day rollover, and deficits change the projection.
Limitations
It does not guarantee the user will study; it only compares planned demand with declared capacity and existing contracts.
Decision it supports
Decide whether to reduce new-card intake, recover deficit, or redistribute study.
Version and changelog
Public contract derived from `metric-policy-v1` and the Gate 1 canonical metric envelope. Future changes must update the catalog before publishing a new interpretation.