AXN as Anti-Suppression Infrastructure (v5 — the text, seated), deposit #1606 of the Crimson Hexagonal Archive (10 September 2026), carries the ~15,300-word paper that deposit #1095 declared on 18 July 2026 and never held.
#1095 was minted as the source-engaged supersession of #1094 and its canonical text was never written. What stood at its text path was 2,147 characters of frontmatter, title and abstract — 135 prose words after scaffold removal. The archive detected this: #1095's body_status carries recovery_status PASS_COMPLETE_TEXT_NOT_SEATED, audited 31 July, with a note recording that the earlier label COMPLETE had denoted pass completion and been read as text completeness, and was renamed so the two claims could not be confused. Ninety-eight deposits carry that status.
This is a new deposit rather than a repair because the AXN is content-derived and #1095's glyphs were minted from the stub: the sha256 of the 2,147-byte file derives AXN:0458.EMPIRICAL.🌸♠️🪦📎⚖️🚀 exactly, while the full text derives a different sequence. AXN:0458 is therefore a true content-derived name for the wrong content, and writing the paper into that record would break the correspondence the identifier system exists to guarantee. #1095 stands as the valid name of its byte-state — the same rule under which it superseded #1094.
The paper argues that AXN belongs to a lineage of anti-suppression infrastructure rather than being novel. Its second section surveys seven cases of survival under all-over suppression — Poland's Second Circulation, Soviet samizdat, pirate radio, the abolitionist press and mirror topology, cypherpunk protocol-as-armor, OONI, and emergency preservation in Data Refuge, SUCHO and LOCKSS — and extracts aftermath documentation as the cross-cutting property before stating a formula. Sections II and III were composed under the rule that engagement with sources is composition-primary: reading the literature is part of writing the paper rather than a stage preceding it.
Appendix D reproduces the Round 1 loss notice verbatim, recording that one INKLING contribution and the research sweep were never recovered after a working session lost its toolset before commit. That the paper about surviving suppression carries its own loss notice as an appendix, and then went six weeks without its text, is not remarked on in the paper.