docs(index): OtOpcUa hw4 batch #611/#592 history entry

Claude-Session: https://claude.ai/code/session_01Ef2ozxS7QPanGT8uoATbTe

.claude/skills/sister-project-history/otopcua.md

@@ -125,3 +125,14 @@ AddRoute of the dev Mac (`oee-tagdump-mac`, `10.181.206.245.1.1`) on the two tes
   served only by the HistorianGateway backend, client-visible results unchanged; a stored resilience config with a `HistoryRead` key
   still parses. Client.CLI/UI now surface a Bad HistoryRead status instead of "0 values returned". Not ScadaBridge-affecting: #654
   central's known-cluster PSK gate retries every 2 s (≤15) while unpublished instead of staying fail-open up to 60 s after boot.
+
+- **2026-10-09 — hw4 batch (#611 / #592, AbCip/native AB transport), local master `6c6800e5` (unpushed at time of writing).**
+  (1) **First native AbCip writes on real hardware**: Z35162 (CompactLogix 5069-L320ERS2) — DINT/BOOL/REAL scratch tags
+  (logic-unreferenced per an L5X census, absent from the Galaxy map) written, read back, restored; user-approved one-off.
+  Read-only press evidence: a mixed STRING/array/scalar batch rides one Multiple Service Packet (#582); symbol listing pages
+  ascend (#572). (2) **AbCip ALMA projection** added; **ALMD fixed to poll `InAlarm`** (it polled `InFaulted`, the input-health
+  flag — never live-gated, so no field impact). New shared-projector rule: a flagged `Unspecified` moves message/severity
+  with no `/alerts` row. (3) Browse-discovered **UDT member writes** work (the "deferred" note was stale; writable only when
+  authored ReadWrite); whole-UDT writes refused by decision; libplctag writes now refuse a width-mismatched type
+  (`BadTypeMismatch`) instead of writing stale bytes. ScadaBridge-visible: only the libplctag mistyped-write refusal (was a
+  silent Good). Follow-up #658 (OpcUaClient first-sight inactive+acked).