The attack surface

TABLE · SPEC §3.2 Inline table inside the measure: long-exposure (window unbounded, requires any exposed pubkey, a slow machine suffices) against short-exposure (window roughly one block interval, requires a machine fast enough to beat it). Three columns; the distinction matters enormously and almost no coverage separates them.

PROSE · SPEC §3.2 The tractable near-term threat, and the one the tracker measures. What is in it today and how it grows. Source: —— · Dated: ————-——-—— · Review due: ————-——-——

PROSE · SPEC §3.2 Mempool to confirmation. Requires a machine that does not plausibly exist yet — say so, and say what "plausibly" is doing in that sentence. Source: —— · Dated: ————-——-—— · Review due: ————-——-——

CLAIM · SPEC §3.2, §11 Every timeline claim is sourced and dated. State the uncertainty rather than picking a year. Two anchors are already fact-checked and want careful attribution: the roughly-10%-before-2030 figure is best cited as Jaques reporting Gidney rather than as Gidney's own claim (his papers give resource estimates, not probabilities), and Google's move of its internal PQC deadline to 2029 is a secondary-source report. Source: —— · Dated: ————-——-—— · Review due: ————-——-——

PROSE · SPEC §11 A quarterly review is a standing commitment on this page above all others. A stale quantum site is actively misleading. Source: —— · Dated: ————-——-—— · Review due: ————-——-——

PROSE · SPEC §2.1 The bridge back to the tracker: exposure is a measurement of what is visible, not a forecast of what will be spent. Link the number without importing urgency into it. Source: —— · Dated: ————-——-—— · Review due: ————-——-——