Pages
Findings Topline v2

Source: c79 finding-lifecycle — topline bake-off, iteration 2 · 2026-08-26

Iteration on v1's verdict: every element is now a claim — labels are sentences, names are attached, consequences are stated, and a zero says something calm instead of sitting mute. The toggle stresses the world (+30 days) — a topline has to communicate on both kinds of day.

the Meridian Monday, as-is

1 · Briefing lines

Full sentences, one per fact that matters, worst first — tone carries urgency, weight carries importance, every line ends in a name.

Nothing is past SLA — 6 open findings, all inside their targets
3 findings have had no eyes on them — oldest 2 days: ravi's unapproved tool use
Next deadline Sep 4 (18 days): sofia's unapproved tool use
marcus's accepted risk returns Sep 15 — the exception expires and re-enters this queue

2 · Your move, spoken

v1's triage cards, given voices: a headline phrase instead of a category label, the top item named in full, and a calm sentence when the count is zero.

3nobody has looked
Oldest is 2 days unseen — ravi's unapproved tool use
2deadlines this month
Nearest: sofia's unapproved tool use, due Sep 4 (18d)
1accepted, for now
marcus's sandbox required returns Sep 15
6 open findings · 2 high, 4 medium, 0 low · across 4 controls

3 · Annotated runway

Time as position, now with words: zones are labeled with their counts, and every tick is listed by name with its days left. The picture and its caption in one card.

0 overdue
0 due this week
2 due this month
4 on pace
Claude Code · sofia · 18d leftsecrets.yaml · 27d leftapi.telemetry-relay.io · 83d leftCodex · ravi · 87d leftchen-mba · 89d left.env.production · 89d left

4 · Claim cards

Four cards where the label is a verb-shaped claim and the sub-line is the consequence — the stat-card format, but every word earns its place.

6
need remediation
2 high across 4 controls
3
still unseen
oldest 2d — ravi's unapproved tool use
18
days to next deadline
sofia's unapproved tool use, Sep 4
1
return later
marcus's exception expires Sep 15

5 · The paragraph

Maximum verbosity: the topline as composed prose in the report voice — the sentence a CISO would say out loud. Numbers bold, tone inline, nothing to decode.

Meridian has 6 open findings across 4 controls — 2 high, 4 medium, 0 low. Nothing is past SLA. 3 haven't been looked at, the oldest for 2 days. The next deadline is Sep 4 (sofia's unapproved tool use). One accepted risk returns to the queue Sep 15.

6 · Strip, worded

The compact band, but phrases instead of labels — reads left to right as a story: state, attention, time, returns. One line of chrome, no decoding.

6 open · 2H 4M 0L3 unseen — oldest 2dnext deadline Sep 4 (18d)1 returns Sep 15

Composition notes

  • The iteration rule (Mark, 2026-08-26): meaning at a glance. Every element became a claim — labels are sentences, the top item is named, the consequence is stated ("returns to the queue Sep 15"), and zeros speak ("Nothing is past SLA" beats a mute 0). The v1 explainer variants (funnel, grid) were dropped; they taught the model without triaging the day.
  • The stress toggle is part of the evaluation: +30 days crosses sofia's deadline and judges each treatment under pressure. A topline that only works on calm days is decoration.
  • Honest denominators: within-SLA % was dropped from the topline entirely — at 1-judged it's a misleading hero number; it belongs on the SLA report where the window gives it a denominator.
  • Provisional read: 2 (Your move, spoken) for the worklist top — cards for scanability, sentences for meaning; 1 (Briefing lines) if vertical space is cheap; 3's named runway list remains the best "running low" display and could sit beneath either.
  • All derived from the shared world; v1 kept at findings-topline-v1 for comparison. Winner graduates to findings-regroup-v2.

Qpoint Brand Style Guide