docs(remediation): D-55f addendum — the #1030 attribution floor is a measured accident, and the merger arm rests on an interim arrangement

docs/remediation/TASKS.md

@@ -908,6 +908,17 @@ answer. be-coder-08 is charged (by Mos) with an end-to-end throwaway-PR proof in
 that can return ABSENT. Until then, "the field will carry them" is an untested claim and must not be
 load-bearing in any GO.
 
+**Addendum (Mos, 2026-08-05) — the floor is a MEASURED ACCIDENT, not a gate that held.** #1030's
+poster happens to be its plurality author (f10-coder, 6/13), so the floor preserves the
+most-represented contributor — but nobody chose that. f10-coder posted in the ordinary course, long
+before anyone knew posting decided attribution; had a TL, a bot, or whichever seat opened the branch
+posted it, all three authors would be lost and the artifact would look identical. **Recorded per
+Mos's ruling: a favourable accident logged as a success is how the next lane inherits a floor that
+is not there.** The one arm that makes the floor real is the merger, and it is currently satisfied
+only by a temporary arrangement: **Mos is the interim homelab merge executor while `merge-gate`'s
+token lacks `write:repository`** — a temporary arrangement carrying a permanent consequence, written
+down as such.
+
 ### D-55b — two parties quoting a third is not corroboration
 
 The coordinator wrote _"installer-7's live-panel GREEN was the last thing before the command"_ **having