diff --git a/docs/handover-seed-extraction.md b/docs/handover-seed-extraction.md
index 55e6648..0fbf38c 100644
--- a/docs/handover-seed-extraction.md
+++ b/docs/handover-seed-extraction.md
@@ -1,7 +1,24 @@
+> ⚠ **RETIRED — 2026-04-27**
+>
+> This document describes the original TESSERA 4.0 seed extraction
+> workflow using `tessera.db` on the Dell pipeline node. That
+> infrastructure no longer exists.
+>
+> **The current dataset assistant handover is:**
+> [`docs/handover-dataset.md`](handover-dataset.md)
+>
+> This document is preserved as historical record of the seed
+> extraction session that produced the initial 12,005 H9 rows
+> in `data/otivm.sqlite3`. Do not act on any instruction in this
+> document — the Dell, tessera-pipeline CT, and tessera.db are
+> all decommissioned.
+
+---
+
 # Handover — TESSERA 4.0 Seed Extraction
 ### Date: 2026-04-26
-### For: Incoming assistant (dataset development track)
-### Read this completely before doing anything
+### Status: RETIRED — see docs/handover-dataset.md
+### For: Historical reference only
 
 ---
 
@@ -96,8 +113,8 @@ Each H5 contains 343 H9 cells (7 H7 children × 49 H9 per H7).
 Total minimum: 5 × 343 = 1,715 H9 rows.
 
 All five are well within the completed portion of tessera.db —
-the Mediterranean is fully populated. No missing data expected
-except occ_flag (stage 06 not run) and some geo_dep/geo_flag
+the Mediterranean is fully populated through stage 04. No missing data
+expected except occ_flag (stage 06 not run) and some geo_dep/geo_flag
 cells at the edge of MRDS/IGME5000 coverage.
 
 ---