PRODUCTIVITY WAVES — LAID OVER EACH OTHER

Weighted output beats per 2-hour block (a shipped customer build = 3, a delivered/landed/merged/completed ledger beat = 2; status pulses and no-output checks excluded), read from the house's own timestamped ledger lines. Solid = measured. Gaps = NO MEASURE — never interpolated.

00:0002:0004:0006:0008:0010:0012:0014:0016:0018:0020:0022:000306090 Nov 6: NO MEASURE before 18:00 Nov 8: day still open after 02:00

■ Nov 7 — the one full ledger day (385 weighted, 185 beats)   ■ Nov 6 evening (115 weighted from 18:07)   ■ Nov 8 to 01:43 (59 weighted, all in 00:00–02:00)

THE LAST 12 HOURS AS ONE WAVE — Nov 7 13:43 → Nov 8 01:43 (weighted per hour, 248 total, 114 beats)

13141516171819202122230001

DAY-LEVEL DENSITY BEFORE THE LEDGER — PROXY, NOT A WAVE

Oct 24 – Nov 5 logs are claim-extraction records, not timestamped ledgers: hourly shape is NO MEASURE for those days. Bars = documented completion-claims per day file (same completion filter). Comparable as density only — the Nov 3–5 heights ride a much larger logged conversation volume, and Nov 5 was Merry's hospital discharge day: the work ran through it, it did not stop for it.

Oct 246Oct 2521Oct 2625Oct 2736Oct 28112Oct 2942Oct 3083Oct 3142Nov 1108Nov 2114Nov 3376Nov 4651Nov 5466

Method + findings: Apex Papers, “Productivity Waves.” Counts verified by spot-check against the raw logs (Nov 6 / Nov 7 / Nov 8).