The sparkline colours by value (kefc7)
The fleet's throughput sparkline maps each point to one of eight block levels. That level — how hot the point is — was computed and then thrown away, and the whole line rendered in a single colour. It now drives a ramp from the muted tone at the bottom of the series to the caller's colour at the top, through the same shared ramp helper the stall indicator uses.
Two constraints shape it. Equal-colour points are emitted as RUNS rather than one node per character: a sparkline renders once per live run in the fleet list, so a node per cell would multiply the node count in the hottest list in the app — a flat series collapses to one node. And the left pad stands for ABSENT data, so it takes the rule tone rather than a series colour; colouring the series while leaving the pad on the low end would have made missing readings look like real low ones.
npm run selftest:tui-sparkline-heat renders the real component and reads the runs back out of its
element tree, asserting the width is unchanged at every point count and that the node count grows
with bands rather than cells.
Known Limitations and Deferred Work. In the ansi variant the palette's tones are
terminal-defined names rather than hex, so the ramp cannot interpolate and collapses to two tones —
the shape still carries the trend there, the intensity does not.
Meter tracks belong to their own meter (kw1e9)
Two leftovers from the sub-cell meter work.
The accent budget — two claims per screen (kpbk7)
The TUI carried 282 ACCENT references and a single hub screen painted twenty-plus orange spans: the header border and its breadcrumb, every key in the footer hint bar, each folder header, the "N...