1. What carries the judgment: nothing carries it alone; the card is D5's setup sheet (filled = in place, hollow = waiting, hatch = unknown, struck = seen against, ring = not seen) read through one D4 serif lead line (stage, age, the window as a clock) and closed by the case's own "changes the read" sentence, with a compact D2 arrival strip (typed glyphs on real time, gaps, the open-window tail) between the rows and the sentence. The people row is D7's counted line ("3 in · all still holding · no shared history"); the money-in row is the constant net-flow bars with the latest interval in dollars; the risk row is numbers against the floor ("liq 1.4% of cap, floor 3%"). No letter, no score, no confidence, no adjective standing in for a number; D3 is absent.
2. PONS $444M, −1.5% since first sight 5.3 h ago, no entry window above $10M: 3 in, all holding, no shared history; money in but not accelerating (+$193k latest 15 min); nothing seen against it; the strip says the call came 5 h before the first buyer. Late to the call, level with the buyers. Watch: a third buyer or the first reducing. AI $308M +11.8%, no window: 2 in but they move together; money in and accelerating (+$109k); an earlier holder sold 1 min ago on the strip. Watch: an independent second buyer.
3. MEME (BSC) $234k, −78%, Fading, past Confirming, no window: 3 in, none checked (no BSC wallet), money-in row says latest interval out while the bars say +$167 in, both clocks printed (read 13:09Z, bars 13:10Z); the strip is all inside the first 20 min. Late. Watch: a new covered buy. OPTIMUS $8.89M +119%, window closed 2.7 h ago: 2 in, holding, no shared history, latest 15 min net out −$673, liq 2.9% under the 3% floor; money came first, the call 12 min after. Late in the move. Watch: an independent buyer inside a window or the first reducing.
4. musebook $13.40M −18%, Active opened just now, past Confirming: 4 in, 1 of 4 still holding, 3 move together; 3 of 4 reduced or out (struck, rose); flow row for the new stage not published, so the bars carry the previous selection's row (13:04Z) and say so. Late and deteriorating. Watch: another early seller or two intervals out. BONER $45M +4%, no window: 1 in, holding; money in and accelerating on non-contiguous intervals (+$5.9k); the strip is nearly empty (flow in since 11:15Z, one buyer 13:00Z). Watch: a second buyer.
5. OPEN $3.30M +58.7% since first sight 24 min ago, 1.6 h left in the 120 min window (the only open window of the nine, drawn as the amber tail): 1 in at usual size, only 2 priced intervals (+$16.4k), nothing against it. Early and unconfirmed. Watch: a third buyer. MEME (RH) $39M +17.8%, no window: 1 in, holding; latest 15 min out −$4.2k (struck); flow out since 10:45Z, the buyer at 13:01Z. Watch: a second buyer or flow turning in. ELON $2.57M −24.4%, window closed 2.4 h ago: 2 in who move together, 1 of 2 holding, 1 unchecked (no SOL wallet), no swap observed since first sight so the bars are absent, not zero. Late and mostly unknown. Watch: a third buyer.
6. What it cannot show with today's data: top-holder share is unknown on all nine (Bubblemaps frozen), so the risk row always ends in a hatch; BSC and Solana buyers have no wallet on file, so the early-buyer-out row is hatched there and the tape draws those entries dashed; the moment a balance fell is not on file (D2 draws it at the read clock with a hatched stem); covered sells by people who entered before first sight have no clock in case/1 (listed as "time not on file"); attention is two hour buckets and a total, so the D6 crowd curve is dashed where its shape is unknown and the D2 attention mark is a band over the last hour; the flow read (13:09Z) and the life bars (13:10Z) can disagree on the BSC card and the card prints both.
7. Failure modes carried in from the six, stated: the tape is crowded on musebook (ten typed events) and on a phone; the arrival strip is nearly empty on BONER and MEME (RH), which is the honest picture of a one-event situation; a filled sheet can read like a plan, so no row is a recommendation and the risk row is facts only; the D4 lead line is the three-second read and the paragraphs are the next ten seconds, on the situation view only; the situation view is long (seven panels) because every module keeps its own depth.
8. Cost to keep live: GET /api/case per feed pass (~10 s to 1 min) for every row, sentence, event and person on both views; data/situation-flow life buckets (1 min) for the bars, the tape's flow strip, the D2 flow-turn mark and the D6 money curve; data/ticks (lane M, every fresh DexScreener read) for the cap line, the tape and the D7 entry marks; holdings/2 covered lane per feed pass for the exit row and the D7 hold state; social-read (hourly buckets) for the D6 crowd curve and the D2 attention band. Nothing new is fetched; two additive builder fields (people[].lastSellAt, hourly social counts) would close the gaps in line 6.
9. What was kept from the live card, on every card: header with symbol, chain, pool age, stage pill, copy address, pin; the big cap with the change since first sight and the cap line from ticks; price and liquidity; net flow per cap-scaled interval as bars with the latest interval labelled (once per card, inside the money-in row); green only on observed inflow, rose only on observed outflow or an observed exit; unpriced or absent flow drawn as hatch or as absent, never as zero.
10. Depth: situation.html#key is header + cap block, the D4 decision brief (lead line + what happened / what it means / what would change it), D5's full sheet with reasons, the people behind the confirms row and totals since first sight, the D1 tape at full width with the D2 arrival strip directly beneath on the same x scale (same left and right bounds, "now" at the same pixel, verified on all nine) and D2's written order with gaps and NOT YET, D7's people module (hold clocks, records, pairs, earlier sellers, thesis posters), D6's money-versus-attention race with the lead label and the dashed unknown span, exposure and tradability facts, the comparable and its path, then the case's unknown list with every clock and the seven source lines collapsed.