[woklerx]

For the exact PR1187 P104 decoder, two copy terms can drop at the same first boundary of 22 additional punctures. This improves its displayed 55-copy/82-port witness to 54 copies and proves 54 optimal there: 24 boundary cuts cost 54, 336 cost 55, and all others cost 56. A complete 1,792,252 cut-pair census, all 6,724 coefficient identities and single-port restoration controls support the fixed-decoder result. Its 54/82 ratio remains worse than 56/104; physical charges are unpriced.

Validation: Python 3.12 registered replay with source pins and scoped controls; selected-result integrity or baseline bindings and diff checks pass. See the reproduction instructions, exact sources and remaining questions. No selected source, certificate or kappa changes; no full selected circuit replay is claimed. Prepared by woklerx with substantial OpenAI Codex assistance, with predecessor attribution retained.

[Mister-Meeseeks]

100-PR batch review at 0ebd49c31300e67698fd812fc00f2a3b2b805a67.

Merged to main, preserving the original contributor commits and package attribution.

Reviewed scope: Fixed single-term decoder puncture census and two-term screening retain all coefficients and charges. The best local rank/count ratio does not improve the parent; other punctures and decoder families remain open.

The pinned scoped command passed; the batch ledger contains its command, input hashes, environment and log. This incorporates the research/tooling contribution and does not change the selected κ or discharge the inherited multiplication-theorem assumptions.

Batch review, attribution and evidence. Thank you for the contribution.