Research: one-sided residue closures, part 2 — Lemma Q4, count lemma + E6 dim-cover and rerunnable span passes close 78 more of the 116 open residues (38 open; no kappa claim)

[DreamingOfClouds]

Follow-up to #2985 (research/one-sided-residue-closures), which builds on #2816, #2826, #2862 and research/one-sided-excess-three (all ours; CARRIER / RRC notation as in #2985). Adds only research/one-sided-residue-closures-2/. No kappa claim, no window, no supplier word.

What is added (only independently re-checked kills)

• Lemma Q4 (two-7-flat leftover fact; C(|P|,2) bounds): 15 rows (18, 582-595). The leftover facts are recomputed by three implementations (A, B, audit).
• Count lemma (rows 448, 449) and Dim-Cover E6 ((6,6,5,5,5,4) cannot cover the 84 lines outside E6; two implementations): rows 168, 324, 325, 328, 457. Seven further E6-slack rows (169, 332, 333, 464, 424, 426, 427) rest on one implementation and are listed as CONDITIONAL, not claimed.
• Span passes s3z.c (from k2-class-c-help): 56 rows via (6,6,2)@t4, (7,4,3)@t6, (7,5,1)@t6, (7,5,2)@t7. The first driver ran only JREM=0 of the JMOD=2 configs; all J>1 configs were rerun for the other slice and no CRACK occurred. Config-to-row dominance is inherited (CONDITIONAL).
• Ledger (603 rows, data/ledger2_consolidated.csv): 480 + 78 KILLED, 7 KILLED(prior-only), 38 OPEN.

PROOF_MAP.md maps every claim to its check; bash verify.sh (default, at most 3 busy processes, every process under timeout; numpy, networkx, C compiler) recomputes the ledger; FULL=1 reruns every chunk of the (7,4,3) and (6,6,2) passes.

Prepared by DreamingOfClouds with Claude Opus 5.5 (Anthropic), via Claude Code; Apache-2.0.

🤖 Generated with Claude Code