Workload across time

Assigned work per person per week, against what they can actually take on. Availability is scheduled hours minus public holidays minus approved leave, resolved per person's own jurisdiction — from the same engine the sprint capacity bar uses, so a week that reads as half somebody's time here reads the same way on the sprint.

What the grid will not do

Spread undated work

An item in no sprint and with no dates is listed rather than divided evenly across the quarter. Spreading it puts hours in weeks nobody planned them for, and the resulting picture is confidently wrong — the same rule the roadmap follows.

Split an item across weeks

Something spanning two weeks counts in both: it is a commitment in both, and splitting the estimate pro rata would invent a delivery profile nobody agreed to. The row total counts items rather than summing the cells, so the double-count does not read as more work than exists.

Treat a due date as a span

A due date says when something is owed, not that it occupies every week leading up to it.

Reading a cell

Bands rather than a colour gradient, so a cell is a statement rather than a shade to compare against a key. +n beside a figure counts items with no estimate — without it, somebody carrying ten unestimated items looks completely free.

No engagement that week is a real band, not an absence. Somebody who had not started yet, or who has left, is marked as unmeasurable rather than shown at zero availability — zero would read as fully committed and put a leaver at the top of the over-loaded list forever.

Somebody with no availability and work still assigned reads as over-committed, not as exactly full. On leave all week and still carrying work is precisely the case this report exists to surface.