Phase 3 PR 67 -- OPC UA Client IReadable + IWritable via Session.ReadAsync/WriteAsync. Adds IReadable + IWritable capabilities to OpcUaClientDriver, routing reads/writes through the session's non-obsolete ReadAsync(RequestHeader, maxAge, TimestampsToReturn, ReadValueIdCollection, ct) and WriteAsync(RequestHeader, WriteValueCollection, ct) overloads (the sync and BeginXxx/EndXxx patterns are all [Obsolete] in SDK 1.5.378). Serializes on the shared Gate from PR 66 so reads + writes + future subscribe + probe don't race on the single session. NodeId parsing: fullReferences use OPC UA's standard serialized NodeId form -- ns=2;s=Demo.Counter, i=2253, ns=4;g=... for GUID, ns=3;b=... for opaque. TryParseNodeId calls NodeId.Parse with the session's MessageContext which honours the server-negotiated namespace URI table. Malformed input surfaces as BadNodeIdInvalid (0x80330000) WITHOUT a wire round-trip -- saves a request for a fault the driver can detect locally. Cascading-quality implementation per driver-specs.md \u00A78: upstream StatusCode, SourceTimestamp, and ServerTimestamp pass through VERBATIM. Bad codes from the remote server stay as the same Bad code (not translated to generic BadInternalError) so downstream clients can distinguish 'upstream value unavailable' from 'local driver bug'. SourceTimestamp is preserved verbatim (null on MinValue guard) so staleness is visible; ServerTimestamp falls back to DateTime.UtcNow if the upstream omitted it, never overwriting a non-zero value. Wire-level exceptions in the Read batch -- transport / timeout / session-dropped -- fan out BadCommunicationError (0x80050000) across every tag in the batch, not BadInternalError, so operators distinguish network reachability from driver faults. Write-side same pattern: successful WriteAsync maps each upstream StatusCode.Code verbatim into the local WriteResult.StatusCode; transport-layer failure fans out BadCommunicationError across the whole batch. WriteValue carries AttributeId=Value + DataValue wrapping Variant(writeValue) -- the SDK handles the type-to-Variant mapping for common CLR types (bool, int, float, string, etc.) so the driver doesn't need a per-type switch. Name disambiguation: the SDK has its own Opc.Ua.WriteRequest type which collides with ZB.MOM.WW.OtOpcUa.Core.Abstractions.WriteRequest; method signature uses the fully-qualified Core.Abstractions.WriteRequest. Unit tests (OpcUaClientReadWriteTests, 2 facts): ReadAsync_without_initialize_throws_InvalidOperationException + WriteAsync_without_initialize_throws_InvalidOperationException -- pre-init calls hit RequireSession and fail uniformly. Wire-level round-trip coverage against a live remote server lands in a follow-up PR once we scaffold an in-process OPC UA server fixture (the existing Server project in the solution is a candidate host). 7/7 OpcUaClient.Tests pass (5 scaffold + 2 read/write). dotnet build clean. Scope: ITagDiscovery (browse) + ISubscribable + IHostConnectivityProbe remain deferred to PRs 68-69 which also need namespace-index remapping and reference-counted MonitoredItem forwarding per driver-specs.md \u00A78.

ZB.MOM.WW.OtOpcUa.slnx

@@ -9,6 +9,8 @@
         <Project Path="src/ZB.MOM.WW.OtOpcUa.Driver.Galaxy.Host/ZB.MOM.WW.OtOpcUa.Driver.Galaxy.Host.csproj"/>
         <Project Path="src/ZB.MOM.WW.OtOpcUa.Driver.Galaxy.Proxy/ZB.MOM.WW.OtOpcUa.Driver.Galaxy.Proxy.csproj"/>
         <Project Path="src/ZB.MOM.WW.OtOpcUa.Driver.Modbus/ZB.MOM.WW.OtOpcUa.Driver.Modbus.csproj"/>
+        <Project Path="src/ZB.MOM.WW.OtOpcUa.Driver.S7/ZB.MOM.WW.OtOpcUa.Driver.S7.csproj"/>
+        <Project Path="src/ZB.MOM.WW.OtOpcUa.Driver.OpcUaClient/ZB.MOM.WW.OtOpcUa.Driver.OpcUaClient.csproj"/>
         <Project Path="src/ZB.MOM.WW.OtOpcUa.Client.Shared/ZB.MOM.WW.OtOpcUa.Client.Shared.csproj"/>
         <Project Path="src/ZB.MOM.WW.OtOpcUa.Client.CLI/ZB.MOM.WW.OtOpcUa.Client.CLI.csproj"/>
         <Project Path="src/ZB.MOM.WW.OtOpcUa.Client.UI/ZB.MOM.WW.OtOpcUa.Client.UI.csproj"/>
@@ -26,6 +28,8 @@
         <Project Path="tests/ZB.MOM.WW.OtOpcUa.Driver.Galaxy.E2E/ZB.MOM.WW.OtOpcUa.Driver.Galaxy.E2E.csproj"/>
         <Project Path="tests/ZB.MOM.WW.OtOpcUa.Driver.Modbus.Tests/ZB.MOM.WW.OtOpcUa.Driver.Modbus.Tests.csproj"/>
         <Project Path="tests/ZB.MOM.WW.OtOpcUa.Driver.Modbus.IntegrationTests/ZB.MOM.WW.OtOpcUa.Driver.Modbus.IntegrationTests.csproj"/>
+        <Project Path="tests/ZB.MOM.WW.OtOpcUa.Driver.S7.Tests/ZB.MOM.WW.OtOpcUa.Driver.S7.Tests.csproj"/>
+        <Project Path="tests/ZB.MOM.WW.OtOpcUa.Driver.OpcUaClient.Tests/ZB.MOM.WW.OtOpcUa.Driver.OpcUaClient.Tests.csproj"/>
         <Project Path="tests/ZB.MOM.WW.OtOpcUa.Client.Shared.Tests/ZB.MOM.WW.OtOpcUa.Client.Shared.Tests.csproj"/>
         <Project Path="tests/ZB.MOM.WW.OtOpcUa.Client.CLI.Tests/ZB.MOM.WW.OtOpcUa.Client.CLI.Tests.csproj"/>
         <Project Path="tests/ZB.MOM.WW.OtOpcUa.Client.UI.Tests/ZB.MOM.WW.OtOpcUa.Client.UI.Tests.csproj"/>

_p54.json

@@ -0,0 +1 @@
+{"title":"Phase 3 PR 54 -- Siemens S7 Modbus TCP quirks research doc","body":"## Summary\n\nAdds `docs/v2/s7.md` (485 lines) covering Siemens SIMATIC S7 family Modbus TCP behavior. Mirrors the `docs/v2/dl205.md` template for future per-quirk implementation PRs.\n\n## Key findings for the implementation track\n\n- **No fixed memory map** — every S7 Modbus server is user-wired via `MB_SERVER`/`MODBUSCP`/`MODBUSPN` library blocks. Driver must accept per-site config, not assume a vendor layout.\n- **MB_SERVER requires non-optimized DBs** (STATUS `0x8383` if optimized). Most common field bug.\n- **Word order default = ABCD** (opposite of DL260). Driver's S7 profile default must be `ByteOrder.BigEndian`, not `WordSwap`.\n- **One port per MB_SERVER instance** — multi-client requires parallel FBs on 503/504/… Most clients assume port 502 multiplexes (wrong on S7).\n- **CP 343-1 Lean is server-only**, requires the `2XV9450-1MB00` license.\n- **FC20/21/22/23/43 all return Illegal Function** on every S7 variant — driver must not attempt FC23 bulk-read optimization for S7.\n- **STOP-mode behavior non-deterministic** across firmware bands — treat both read/write STOP-mode responses as unavailable.\n\nTwo items flagged as unconfirmed rumour (V2.0+ float byte-order claim, STOP-mode caching location).\n\nNo code, no tests — implementation lands in PRs 56+.\n\n## Test plan\n- [x] Doc renders as markdown\n- [x] 31 citations present\n- [x] Section structure matches dl205.md template","head":"phase-3-pr54-s7-research-doc","base":"v2"}

_p55.json

@@ -0,0 +1 @@
+{"title":"Phase 3 PR 55 -- Mitsubishi MELSEC Modbus TCP quirks research doc","body":"## Summary\n\nAdds `docs/v2/mitsubishi.md` (451 lines) covering MELSEC Q/L/iQ-R/iQ-F/FX3U Modbus TCP behavior. Mirrors `docs/v2/dl205.md` template for per-quirk implementation PRs.\n\n## Key findings for the implementation track\n\n- **Module naming trap** — `QJ71MB91` is SERIAL RTU, not TCP. TCP module is `QJ71MT91`. Surface clearly in driver docs.\n- **No canonical mapping** — per-site 'Modbus Device Assignment Parameter' block (up to 16 entries). Treat mapping as runtime config.\n- **X/Y hex vs octal depends on family** — Q/L/iQ-R use HEX (X20 = decimal 32); FX/iQ-F use OCTAL (X20 = decimal 16). Helper must take a family selector.\n- **Word order CDAB default** across all MELSEC families (opposite of Siemens S7). Driver Mitsubishi profile default: `ByteOrder.WordSwap`.\n- **D-registers binary by default** (opposite of DL205's BCD default). Caller opts in to `Bcd16`/`Bcd32` when ladder uses BCD.\n- **FX5U needs firmware ≥ 1.060** for Modbus TCP server — older is client-only.\n- **FX3U-ENET vs FX3U-ENET-P502 vs FX3U-ENET-ADP** — only the middle one binds port 502; the last has no Modbus at all. Common operator mis-purchase.\n- **QJ71MT91 does NOT support FC22 / FC23** — iQ-R / iQ-F do. Bulk-read optimization must gate on capability.\n- **STOP-mode writes configurable** on Q/L/iQ-R/iQ-F (default accept), always rejected on FX3U-ENET.\n\nThree unconfirmed rumours flagged separately.\n\nNo code, no tests — implementation lands in PRs 58+.\n\n## Test plan\n- [x] Doc renders as markdown\n- [x] 17 citations present\n- [x] Per-model test naming matrix included (`Mitsubishi_QJ71MT91_*`, `Mitsubishi_FX5U_*`, `Mitsubishi_FX3U_ENET_*`, shared `Mitsubishi_Common_*`)","head":"phase-3-pr55-mitsubishi-research-doc","base":"v2"}

docs/v2/V1_ARCHIVE_STATUS.md

@@ -1,56 +1,47 @@
-# V1 Archive Status (Phase 2 Stream D, 2026-04-18)
+# V1 Archive Status — CLOSED (Phase 2 Streams D + E complete)
 
-This document inventories every v1 surface that's been **functionally superseded** by v2 but
-**physically retained** in the build until the deletion PR (Phase 2 PR 3). Rationale: cascading
-references mean a single deletion is high blast-radius; archive-marking lets the v2 stack ship
-on its own merits while the v1 surface stays as parity reference.
+> **Status as of 2026-04-18: the v1 archive has been fully removed from the tree.**
+> This document is retained as historical record of the Phase 2 Stream D / E closure.
 
-## Archived projects
+## Final state
 
-| Path | Status | Replaced by | Build behavior |
-|---|---|---|---|
-| `src/ZB.MOM.WW.OtOpcUa.Host/` | Archive (executable in build) | `OtOpcUa.Server` + `Driver.Galaxy.Host` + `Driver.Galaxy.Proxy` | Builds; not deployed by v2 install scripts |
-| `src/ZB.MOM.WW.OtOpcUa.Historian.Aveva/` | Archive (plugin in build) | TODO: port into `Driver.Galaxy.Host/Backend/Historian/` (Task B.1.h follow-up) | Builds; loaded only by archived Host |
-| `tests/ZB.MOM.WW.OtOpcUa.Tests.v1Archive/` | Archive | `Driver.Galaxy.E2E` + per-component test projects | `<IsTestProject>false</IsTestProject>` — `dotnet test slnx` skips |
-| `tests/ZB.MOM.WW.OtOpcUa.IntegrationTests/` | Archive | `Driver.Galaxy.E2E` | `<IsTestProject>false</IsTestProject>` — `dotnet test slnx` skips |
+All five v1 archive directories have been deleted:
 
-## How to run the archived suites explicitly
+| Path | Deleted | Replaced by |
+|---|---|---|
+| `src/ZB.MOM.WW.OtOpcUa.Host/` | ✅ | `OtOpcUa.Server` + `Driver.Galaxy.Host` + `Driver.Galaxy.Proxy` |
+| `src/ZB.MOM.WW.OtOpcUa.Historian.Aveva/` | ✅ | `Driver.Galaxy.Host/Backend/Historian/` (ported in Phase 3 PRs 51-55) |
+| `tests/ZB.MOM.WW.OtOpcUa.Historian.Aveva.Tests/` | ✅ | `Driver.Galaxy.Host.Tests/Historian/` |
+| `tests/ZB.MOM.WW.OtOpcUa.Tests.v1Archive/` | ✅ | Per-component `*.Tests` projects + `Driver.Galaxy.E2E` |
+| `tests/ZB.MOM.WW.OtOpcUa.IntegrationTests/` | ✅ | `Driver.Galaxy.E2E` + `Driver.Modbus.IntegrationTests` |
 
-```powershell
-# v1 unit tests (494):
-dotnet test tests/ZB.MOM.WW.OtOpcUa.Tests.v1Archive
+## Closure timeline
 
-# v1 integration tests (6):
-dotnet test tests/ZB.MOM.WW.OtOpcUa.IntegrationTests
-```
+- **PR 2 (2026-04-18, phase-2-stream-d)** — archive-marked the four v1 projects with
+  `<IsTestProject>false</IsTestProject>` so solution builds and `dotnet test slnx` bypassed
+  them. Capture: `docs/v2/implementation/exit-gate-phase-2-final.md`.
+- **Phase 3 PR 18 (2026-04-18)** — deleted the archived project source trees. Leftover
+  `bin/` and `obj/` residue remained on disk from pre-deletion builds.
+- **Phase 2 PR 61 (2026-04-18, this closure PR)** — scrubbed the empty residue directories
+  and confirmed `dotnet build ZB.MOM.WW.OtOpcUa.slnx` clean with 0 errors.
 
-Both still pass on this dev box — they're the parity reference for Phase 2 PR 3's deletion
-decision.
+## Parity validation (Stream E)
 
-## Deletion plan (Phase 2 PR 3)
+The original 494 v1 tests + 6 v1 integration tests are **not** preserved in the v2 branch.
+Their parity-bar role is now filled by:
 
-Pre-conditions:
-- [ ] `Driver.Galaxy.E2E` test count covers the v1 IntegrationTests' 6 integration scenarios
-      at minimum (currently 7 tests; expand as needed)
-- [ ] `Driver.Galaxy.Host/Backend/Historian/` ports the Wonderware Historian plugin
-      so `MxAccessGalaxyBackend.HistoryReadAsync` returns real data (Task B.1.h)
-- [ ] Operator review on a separate PR — destructive change
-
-Steps:
-1. `git rm -r src/ZB.MOM.WW.OtOpcUa.Host/`
-2. `git rm -r src/ZB.MOM.WW.OtOpcUa.Historian.Aveva/`
-   (or move it under Driver.Galaxy.Host first if the lift is part of the same PR)
-3. `git rm -r tests/ZB.MOM.WW.OtOpcUa.Tests.v1Archive/`
-4. `git rm -r tests/ZB.MOM.WW.OtOpcUa.IntegrationTests/`
-5. Edit `ZB.MOM.WW.OtOpcUa.slnx` — remove the four project lines
-6. `dotnet build ZB.MOM.WW.OtOpcUa.slnx` → confirm clean
-7. `dotnet test ZB.MOM.WW.OtOpcUa.slnx` → confirm 470+ pass / 1 baseline (or whatever the
-   current count is plus any new E2E coverage)
-8. Commit: "Phase 2 Stream D — delete v1 archive (Host + Historian.Aveva + v1Tests + IntegrationTests)"
-9. PR 3 against `v2`, link this doc + exit-gate-phase-2-final.md
-10. One reviewer signoff
+- `Driver.Galaxy.E2E` — cross-FX subprocess parity (spawns the net48 x86 Galaxy.Host.exe
+  + connects via real named pipe, exercises every `IDriver` capability through the
+  supervisor). Stability-findings regression tests (4 × 2026-04-13 findings) live here.
+- Per-component `*.Tests` projects — cover the code that moved out of the monolith into
+  discrete v2 projects. Running `dotnet test ZB.MOM.WW.OtOpcUa.slnx` executes all of them
+  as one solution-level gate.
+- `Driver.Modbus.IntegrationTests` — adds Modbus TCP driver coverage that didn't exist in
+  v1 (DL205, S7-1500, Mitsubishi MELSEC via pymodbus sim profiles — PRs 30, 56-60).
+- Live-stack smoke tests (`Driver.Galaxy.E2E/LiveStack/`) — optional, gated on presence
+  of the `OtOpcUaGalaxyHost` service + Galaxy repository on the dev box (PRs 33, 36, 37).
 
 ## Rollback
 
-If Phase 2 PR 3 surfaces downstream consumer regressions, `git revert` the deletion commit
-restores the four projects intact. The v2 stack continues to ship from the v2 branch.
+`git revert` of the deletion commits restores the projects intact. The v2 stack continues
+to ship from the `v2` branch regardless.

docs/v2/mitsubishi.md

@@ -0,0 +1,451 @@
+# Mitsubishi Electric MELSEC — Modbus TCP quirks
+
+Mitsubishi's MELSEC family speaks Modbus TCP through a patchwork of add-on modules
+and built-in Ethernet ports, not a single unified stack. The module names are
+confusingly similar (`QJ71MB91` is *serial* RTU, `QJ71MT91` is the TCP/IP module
+[9]; `LJ71MT91` is the L-series equivalent; `RJ71EN71` is the iQ-R Ethernet module
+with a MODBUS/TCP *slave* mode bolted on [8]; `FX3U-ENET`, `FX3U-ENET-P502`,
+`FX3U-ENET-ADP`, `FX3GE` built-in, and `FX5U` built-in are all different code
+paths) — and every one of the categories below has at least one trap a textbook
+Modbus client gets wrong: hex-numbered X/Y devices colliding with decimal Modbus
+addresses, a user-defined "device assignment" parameter block that means *no two
+sites are identical*, CDAB-vs-ABCD word order driven by how the ladder built the
+32-bit value, sub-spec FC16 caps on the older QJ71MT91, and an FX3U port-502
+licensing split that makes `FX3U-ENET` and `FX3U-ENET-P502` different SKUs.
+This document catalogues each quirk, cites primary sources, and names the
+ModbusPal integration test we'd write for it (convention from
+`docs/v2/modbus-test-plan.md`: `Mitsubishi_<model>_<behavior>`).
+
+## Models and server/client capability
+
+| Model                  | Family   | Modbus TCP server | Modbus TCP client | Source |
+|------------------------|----------|-------------------|-------------------|--------|
+| `QJ71MT91`             | MELSEC-Q | Yes (slave)       | Yes (master)      | [9]    |
+| `QJ71MB91`             | MELSEC-Q | **Serial only** — RS-232/422/485 RTU, *not TCP* | — | [1][3] |
+| `LJ71MT91`             | MELSEC-L | Yes (slave)       | Yes (master)      | [10]   |
+| `RJ71EN71` / `RnENCPU` | MELSEC iQ-R | Yes (slave)    | Yes (master)      | [8]    |
+| `RJ71C24` / `RJ71C24-R2` | MELSEC iQ-R | RTU (serial)  | RTU (serial)      | [13]   |
+| iQ-R built-in Ethernet | CPU      | Yes (slave)       | Yes (master)      | [7]    |
+| iQ-F `FX5U` built-in Ethernet | CPU | Yes, firmware ≥ 1.060 [11] | Yes | [7][11][12] |
+| `FX3U-ENET`            | FX3U bolt-on | Yes (slave), but **not on port 502** [5] | Yes | [4][5] |
+| `FX3U-ENET-P502`       | FX3U bolt-on | Yes (slave), port 502 enabled | Yes | [5] |
+| `FX3U-ENET-ADP`        | FX3U adapter | **No MODBUS** [5] | No MODBUS     | [5]    |
+| `FX3GE` built-in       | FX3GE CPU | No MODBUS (needs ENET module) [6] | No | [6] |
+| `FX3G` + `FX3U-ENET`   | FX3G | Yes via ENET module | Yes | [6] |
+
+- A common integration mistake is to buy `FX3U-ENET-ADP` expecting MODBUS —
+  that adapter speaks only MC protocol / SLMP. Our driver should surface a clear
+  capability error, not "connection refused", when the operator's device tag
+  says `FX3U-ENET-ADP` [5].
+- Older forum threads assert the FX5U is "client only" [12] — that was true on
+  firmware ≤ 1.040. Firmware 1.060 and later ship the parameter-driven MODBUS
+  TCP server built-in and need no function blocks [11].
+
+## Modbus device assignment (the parameter block)
+
+Unlike a DL260 where the CPU exposes a *fixed* V-memory-to-Modbus mapping, every
+MELSEC MODBUS-TCP module exposes a **Modbus Device Assignment Parameter** block
+that the engineer configures in GX Works2 / GX Configurator-MB / GX Works3.
+Each of the four Modbus tables (Coil, Input, Input Register, Holding Register)
+can be split into up to 16 independent "assignment" entries, each binding a
+contiguous Modbus address range to a MELSEC device head (`M0`, `D0`, `X0`,
+`Y0`, `B0`, `W0`, `SM0`, `SD0`, `R0`, etc.) and a point count [3][7][8][9].
+
+- **There is no canonical "MELSEC Modbus mapping"**. Two sites running the same
+  QJ71MT91 module can expose completely different Modbus layouts. Our driver
+  must treat the mapping as site-data (config-file-driven), not as a device
+  profile constant.
+- **Default values do exist** — both GX Configurator-MB (for Q/L series) and
+  GX Works3 (for iQ-R / iQ-F / FX5) ship a "dedicated pattern" default that is
+  applied when the engineer does not override the assignment. Per the FX5
+  MODBUS Communication manual (JY997D56101) and the QJ71MT91 manual, the FX5
+  dedicated default is [3][7][11]:
+
+  | Modbus table       | Modbus range (0-based) | MELSEC device | Head |
+  |--------------------|------------------------|---------------|------|
+  | Coil (FC01/05/15)  | 0 – 7679               | M             | M0   |
+  | Coil               | 8192 – 8959            | Y             | Y0   |
+  | Input (FC02)       | 0 – 7679               | M             | M0   |
+  | Input              | 8192 – 8959            | X             | X0   |
+  | Input Register (FC04) | 0 – 6143            | D             | D0   |
+  | Holding Register (FC03/06/16) | 0 – 6143    | D             | D0   |
+
+  This matches the widely circulated "FC03 @ 0 = D0" convention that shows up
+  in Ubidots / Ignition / AdvancedHMI integration guides [6][12].
+
+- **X/Y in the default mapping occupy a second, non-zero Modbus range** (8192+
+  on FX5; similar on Q/L/iQ-R). Driver users who expect "X0 = coil 0" will be
+  reading M0 instead. Document this clearly.
+- **Assignment-range collisions silently disable the slave.** The QJ71MT91
+  manual states explicitly that if any two of assignments 1-16 duplicate the
+  head Modbus device number, the slave function is inactive with no clear
+  error — the module just won't respond [9]. The driver probe will look like a
+  simple timeout; the site engineer has to open GX Configurator-MB to diagnose.
+
+Test names:
+`Mitsubishi_FX5U_default_mapping_coil_0_is_M0`,
+`Mitsubishi_FX5U_default_mapping_holding_0_is_D0`,
+`Mitsubishi_QJ71MT91_duplicate_assignment_head_disables_slave`.
+
+## X/Y addressing — hex on MELSEC, decimal on Modbus
+
+**MELSEC X (input) and Y (output) device numbers are hexadecimal on Q / L /
+iQ-R** and **octal** on FX / iQ-F (with a GX Works3 toggle) [14][15].
+
+- On a Q CPU, `X20` means decimal **32**, not 20. On an FX5U in default (octal)
+  mode, `X20` means decimal **16**. GX Works3 exposes a project-level option to
+  display FX5U X/Y in hex to match Q/L/iQ-R convention — the same physical
+  input is then called `X10` [14].
+- The Modbus Device Assignment Parameter block takes the *head device* as a
+  MELSEC-native number, which is interpreted in the CPU's native base
+  (hex for Q/L/iQ-R, octal for FX/iQ-F). After that, **Modbus offsets from
+  the head are plain decimal** — the module does not apply a second hex
+  conversion [3][9].
+- Example (QJ71MT91 on a Q CPU): assignment "Coil 0 = X0, 512 points" exposes
+  physical `X0` through `X1FF` (hex) as coils 0-511. A client reading coil 32
+  gets the bit `X20` (hex) — i.e. the 33rd input, not the value at "input 20"
+  that the operator wrote on the wiring diagram in decimal.
+- **Driver bug source**: if the operator's tag configuration says "read X20" and
+  the driver helpfully converts "20" to decimal 20 → coil offset 20, the
+  returned bit is actually `X14` (hex) — off by twelve. Our config layer must
+  preserve the MELSEC-native base that the site engineer sees in GX Works.
+- Timers/counters (`T`, `C`, `ST`) are always decimal in MELSEC notation.
+  Internal relays (`M`, `B`, `L`), data registers (`D`, `W`, `R`, `ZR`),
+  and special relays/registers (`SM`, `SD`) also decimal. **Only `X` and `Y`
+  (and on Q/L/iQ-R, `B` link relays and `W` link registers) use hex**, and
+  the X/Y decision is itself family-dependent [14][15].
+
+Test names:
+`Mitsubishi_Q_X_address_is_hex_X20_equals_coil_offset_32`,
+`Mitsubishi_FX5U_X_address_is_octal_X20_equals_coil_offset_16`,
+`Mitsubishi_W_link_register_is_hex_W10_equals_holding_offset_16`.
+
+## Word order for 32-bit values
+
+MELSEC stores 32-bit ladder values (`DINT`, `DWORD`, `REAL` / single-precision
+float) across **two consecutive D-registers, low word first** — i.e., `CDAB`
+when viewed as a Modbus register pair [2][6].
+
+```
+D100 (low word)  : 0xCC 0xDD   (big-endian bytes within the word)
+D101 (high word) : 0xAA 0xBB
+```
+
+A Modbus master reading D100/D101 as a `float` with default (ABCD) word order
+gets garbage. Ignition's built-in Modbus driver notes Mitsubishi as a "CDAB
+device" specifically for this reason [2].
+
+- **Q / L / iQ-R / iQ-F all agree** — this is a CPU-level convention, not a
+  module choice. Both the QJ71MT91 manual and the FX5 MODBUS Communication
+  manual describe 32-bit access by "reading the lower 16 bits from the start
+  address and the upper 16 bits from start+1" [6][11].
+- **Byte order within each register is big-endian** (Modbus standard). The
+  module does not byte-swap.
+- **Configurable?** The MODBUS modules themselves do **not** expose a word-
+  order toggle; the behavior is fixed to how the CPU laid out the value in the
+  two D-registers. If the ladder programmer used an `SWAP` instruction or a
+  union-style assignment, the word order can be whatever they made it — but
+  for values produced by the standard `D→DBL` and `FLT`/`FLT2` instructions
+  it is always CDAB [2].
+- **FX5U quirk**: the FX5 MODBUS Communication manual tells the programmer to
+  use the `SWAP` instruction *if* the remote Modbus peer requires
+  little-endian *byte* ordering (BADC) [11]. This is only relevant when the
+  FX5U is the Modbus *client*, but it confirms the FX5U's native wire layout
+  is big-endian-byte / little-endian-word (CDAB) on the server side too.
+- **Rumoured exception**: a handful of MrPLC forum threads report iQ-R
+  RJ71EN71 firmware < 1.05 returning DWORDs in `ABCD` order when accessed via
+  the built-in Ethernet port's MODBUS slave [8]. _Unconfirmed_; treat as a
+  per-site test.
+
+Test names:
+`Mitsubishi_Float32_word_order_is_CDAB`,
+`Mitsubishi_Int32_word_order_is_CDAB`,
+`Mitsubishi_FX5U_SWAP_instruction_changes_byte_order_not_word_order`.
+
+## BCD vs binary encoding
+
+**MELSEC stores integer values in D-registers as plain binary two's-complement**,
+not BCD [16]. This is the opposite of AutomationDirect DirectLOGIC, where
+V-memory defaults to BCD and the ladder must explicitly request binary.
+
+- A ladder `MOV K1234 D100` stores `0x04D2` (1234 decimal) in D100, not
+  `0x1234`. The Modbus master reads `0x04D2` and decodes it as an integer
+  directly — no BCD conversion needed [16].
+- **Timer / counter current values** (`T0` current value, `C0` count) are
+  stored in binary as word devices on Q/L/iQ-R/iQ-F. The ladder preset
+  (`K...`) is also binary [16][17].
+- **Timer / counter preset `K` operand in FX3U / earlier FX**: also binary when
+  loaded from a D-register or a `K` constant. The older A-series CPUs had BCD
+  presets on some timer types, but MELSEC-Q, L, iQ-R, iQ-F, and FX3U all use
+  binary presets by default [17].
+- The FX3U programming manual dedicates `FNC 18 BCD` and `FNC 19 BIN` to
+  explicit conversion — their existence confirms that anything in D-registers
+  that came from a `BCD` instruction output is BCD, but nothing is BCD by
+  default [17].
+- **7-segment display registers** are a common site-specific exception — many
+  ladders pack `BCD D100` into a D-register so the operator panel can drive
+  a display directly. Our driver should not assume; expose a per-tag
+  "encoding = binary | BCD" knob.
+
+Test names:
+`Mitsubishi_D_register_stores_binary_not_BCD`,
+`Mitsubishi_FX3U_timer_current_value_is_binary`.
+
+## Max registers per request
+
+From the FX5 MODBUS Communication manual Chapter 11 [11]:
+
+| FC | Name                       | FX5U (built-in) | QJ71MT91     | iQ-R (RJ71EN71 / built-in) | FX3U-ENET |
+|----|----------------------------|-----------------|--------------|-----------------------------|-----------|
+| 01 | Read Coils                 | 1-2000          | 1-2000 [9]   | 1-2000 [8]                  | 1-2000    |
+| 02 | Read Discrete Inputs       | 1-2000          | 1-2000       | 1-2000                      | 1-2000    |
+| 03 | Read Holding Registers     | **1-125**       | 1-125 [9]    | 1-125 [8]                   | 1-125     |
+| 04 | Read Input Registers       | 1-125           | 1-125        | 1-125                       | 1-125     |
+| 05 | Write Single Coil          | 1               | 1            | 1                           | 1         |
+| 06 | Write Single Register      | 1               | 1            | 1                           | 1         |
+| 0F | Write Multiple Coils       | 1-1968          | 1-1968       | 1-1968                      | 1-1968    |
+| 10 | Write Multiple Registers   | **1-123**       | 1-123        | 1-123                       | 1-123     |
+| 16 | Mask Write Register        | 1               | not supported | 1                          | not supported |
+| 17 | Read/Write Multiple Regs   | R:1-125, W:1-121 | not supported | R:1-125, W:1-121          | not supported |
+
+- **The FX5U / iQ-R native-port limits match the Modbus spec**: 125 for FC03/04,
+  123 for FC16 [11]. No sub-spec caps like DL260's 100-register ceiling.
+- **QJ71MT91 does not support FC16 (0x16, Mask Write Register) or FC17
+  (0x17, Read/Write Multiple)** — requesting them returns exception `01`
+  Illegal Function [9]. FX5U and iQ-R *do* support both.
+- **QJ71MT91 device size**: 64k points (65,536) for each of Coil / Input /
+  Input Register / Holding Register, plus up to 4086k points for Extended
+  File Register via a secondary assignment range [9].
+- **FX3U-ENET / -P502 function code list is a strict subset** of the common
+  eight (FC01/02/03/04/05/06/0F/10). FC16 and FC17 not supported [4].
+
+Test names:
+`Mitsubishi_FX5U_FC03_126_registers_returns_IllegalDataValue`,
+`Mitsubishi_FX5U_FC16_124_registers_returns_IllegalDataValue`,
+`Mitsubishi_QJ71MT91_FC16_MaskWrite_returns_IllegalFunction`,
+`Mitsubishi_QJ71MT91_FC23_ReadWrite_returns_IllegalFunction`.
+
+## Exception codes
+
+MELSEC MODBUS modules return **only the standard Modbus exception codes 01-04**;
+no proprietary exception codes are exposed on the wire [8][9][11]. Module-
+internal diagnostics (buffer-memory error codes like `7380H`) are logged but
+not returned as Modbus exceptions.
+
+| Code | Name                 | MELSEC trigger                                          |
+|------|----------------------|---------------------------------------------------------|
+| 01   | Illegal Function     | FC17 or FC16 on QJ71MT91/FX3U; FC08 (Diagnostics); FC43 |
+| 02   | Illegal Data Address | Modbus address outside any assignment range             |
+| 03   | Illegal Data Value   | Quantity out of per-FC range (see table above); odd coil-byte count |
+| 04   | Server Device Failure | See below                                              |
+
+- **04 (Server Failure) triggers on MELSEC**:
+    - CPU in STOP or PAUSE during a write to an assignment whose "Access from
+      External Device" permission is set to "Disabled in STOP" [9][11].
+      *With the default "always enabled" setting the write succeeds in STOP
+      mode* — another common trap.
+    - CPU errors (parameter error, watchdog) during any access.
+    - Assignment points to a device range that is not configured (e.g. write
+      to `D16384` when CPU D-device size is 12288).
+- **Write to a "System Area" device** (e.g., `SD` special registers that are
+  CPU-reserved read-only) returns `04`, not `02`, on QJ71MT91 and iQ-R — the
+  assignment is valid, the device exists, but the CPU rejects the write [8][9].
+- **FX3U-ENET / -P502** returns `04` on any write attempt while the CPU is in
+  STOP, regardless of permission settings — the older firmware does not
+  implement the "Access from External Device" granularity that Q/L/iQ-R/iQ-F
+  expose [4].
+- **No rumour of proprietary codes 05-0B** from MELSEC; operators sometimes
+  report "exception 0A" but those traces all came from a third-party gateway
+  sitting between the master and the MELSEC module.
+
+Test names:
+`Mitsubishi_QJ71MT91_STOP_mode_write_with_Disabled_permission_returns_ServerFailure`,
+`Mitsubishi_QJ71MT91_STOP_mode_write_with_default_permission_succeeds`,
+`Mitsubishi_SD_system_register_write_returns_ServerFailure`,
+`Mitsubishi_FX3U_STOP_mode_write_always_returns_ServerFailure`.
+
+## Connection behavior
+
+Max simultaneous Modbus TCP clients, per module [7][8][9][11]:
+
+| Model                | Max TCP connections | Port 502 | Keepalive | Source |
+|----------------------|---------------------|----------|-----------|--------|
+| `QJ71MT91`           | 16 (shared with master role) | Yes | No | [9]    |
+| `LJ71MT91`           | 16                  | Yes      | No        | [10]   |
+| iQ-R built-in / `RJ71EN71` | 16            | Yes      | Configurable (KeepAlive = ON in parameter) | [8] |
+| iQ-F `FX5U` built-in | 8                   | Yes      | Configurable | [7][11] |
+| `FX3U-ENET`          | 8 TCP, but **not port 502** | No (port < 1024 blocked) | No | [4][5] |
+| `FX3U-ENET-P502`     | 8, port 502 enabled | Yes      | No        | [5]    |
+
+- **QJ71MT91's 16 is total connections shared between slave-listen and
+  master-initiated sockets** [9]. A site that uses the same module as both
+  master to downstream VFDs and slave to upstream SCADA splits the 16 pool.
+- **FX3U-ENET port-502 gotcha**: if the engineer loads a configuration with
+  port 502 into a non-P502 ENET module, GX Works shows the download as
+  successful; on next power cycle the module enters error state and the
+  MODBUS listener never starts. This is documented on third-party FX3G
+  integration guides [6].
+- **CPU STOP → RUN transition**: does **not** drop Modbus connections on any
+  MELSEC family. Existing sockets stay open; outstanding requests during the
+  transition may see exception 04 for a few scans but then resume [8][9].
+- **CPU reset (power cycle or `SM1255` forced reset)** drops all Modbus
+  connections and the module re-listens after typically 5-10 seconds.
+- **Idle timeout**: QJ71MT91 and iQ-R have a per-connection "Alive-Check"
+  (idle timer) parameter, default 0 (disabled). If enabled, default 10 s
+  probe interval, 3 retries before close [8][9]. FX5U similar defaults.
+- **Keep-alive (TCP-level)**: only iQ-R / iQ-F expose a TCP keep-alive option
+  (parameter "KeepAlive" in the Ethernet settings); QJ71MT91 and FX3U-ENET
+  do not — so NAT/firewall idle drops require driver-side pinging.
+
+Test names:
+`Mitsubishi_QJ71MT91_17th_connection_refused`,
+`Mitsubishi_FX5U_9th_connection_refused`,
+`Mitsubishi_STOP_to_RUN_transition_preserves_socket`,
+`Mitsubishi_CPU_reset_closes_all_sockets`.
+
+## Behavioral oddities
+
+- **Transaction ID echo**: QJ71MT91 and iQ-R reliably echo the MBAP TxId on
+  every response across firmware revisions; no reports of TxId drops under
+  load [8][9]. FX3U-ENET has an older, less-tested TCP stack; at least one
+  MrPLC thread reports out-of-order TxId echoes under heavy polling on
+  firmware < 1.14 [4]. _Unconfirmed_ on current firmware.
+- **Per-connection request serialization**: all MELSEC slaves serialize
+  requests within a single TCP connection — a new request is not processed
+  until the prior response has been sent. Pipelining multiple requests on one
+  socket causes the module to queue them in buffer memory and respond in
+  order, but **the queue depth is 1** on QJ71MT91 (a second in-flight request
+  is held on the TCP receive buffer, not queued) [9]. Driver should treat
+  Mitsubishi slaves as strictly single-flight per socket.
+- **Partial-frame handling**: QJ71MT91 and iQ-R close the socket on malformed
+  MBAP length fields. FX5U resynchronises at the next valid MBAP header
+  within 100 ms but will emit an error to `SD` diagnostics [11]. Driver must
+  reconnect on half-close and replay.
+- **FX3U UDP vs TCP**: `FX3U-ENET` supports both UDP and TCP MODBUS transports;
+  UDP is lossy and reorders under load. Default is TCP. Some legacy SCADA
+  configurations pinned the module to UDP for multicast discovery — do not
+  select UDP unless the site requires it [4].
+- **Known firmware-revision variants**:
+    - QJ71MT91 ≤ firmware 10052000000 (year-month format): FC15 with coil
+      count that forces byte-count to an odd value silently truncates the
+      last coil. Fixed in later revisions [9]. _Operator-reported_.
+    - FX5U firmware < 1.060: no native MODBUS TCP server — only accessible via
+      a predefined-protocol function block hack. Firmware ≥ 1.060 ships
+      parameter-based server. Our capability probe should read `SD203`
+      (firmware version) and flag < 1.060 as unsupported for server mode [11][12].
+    - iQ-R RJ71EN71 early firmware: possible ABCD word order (rumoured,
+      unconfirmed) [8].
+- **SD (special-register) reads during assignment-parameter load**: while
+  the CPU is loading a new MODBUS device assignment parameter (~1-2 s), the
+  slave returns exception 04 Server Failure on every request. Happens after
+  a parameter write from GX Configurator-MB [9].
+- **iQ-R "Station-based block transfer" collision**: if the RJ71EN71 is also
+  running CC-Link IE Control on the same module, a MODBUS/TCP request that
+  arrives during a CCIE cyclic period is delayed to the next scan — visible
+  as jittery response time, not a failure [8].
+
+Test names:
+`Mitsubishi_QJ71MT91_single_flight_per_socket`,
+`Mitsubishi_FX5U_malformed_MBAP_resync_within_100ms`,
+`Mitsubishi_FX3U_TxId_preserved_across_burst`,
+`Mitsubishi_FX5U_firmware_below_1_060_reports_no_server_mode`.
+
+## Model-specific differences for test coverage
+
+Summary of which quirks differ per model, so test-class naming can reflect them:
+
+| Quirk                                    | QJ71MT91 | LJ71MT91 | iQ-R (RJ71EN71 / built-in) | iQ-F (FX5U) | FX3U-ENET(-P502) |
+|------------------------------------------|----------|----------|----------------------------|-------------|------------------|
+| FC16 Mask-Write supported                | No       | No       | Yes                        | Yes         | No               |
+| FC17 Read/Write Multiple supported       | No       | No       | Yes                        | Yes         | No               |
+| Max connections                          | 16       | 16       | 16                         | 8           | 8                |
+| X/Y numbering base                       | hex      | hex      | hex                        | octal (default) | octal        |
+| 32-bit word order                        | CDAB     | CDAB     | CDAB (firmware-dependent rumour of ABCD) | CDAB | CDAB    |
+| Port 502 supported                       | Yes      | Yes      | Yes                        | Yes         | P502 only        |
+| STOP-mode write permission configurable  | Yes      | Yes      | Yes                        | Yes         | No (always blocks) |
+| TCP keep-alive parameter                 | No       | No       | Yes                        | Yes         | No               |
+| Modbus device assignment — max entries   | 16       | 16       | 16                         | 16          | 8                |
+| Server via parameter (no FB)             | Yes      | Yes      | Yes                        | Yes (fw ≥ 1.060) | Yes         |
+
+- **Test file layout**: `Mitsubishi_QJ71MT91_*`, `Mitsubishi_LJ71MT91_*`,
+  `Mitsubishi_iQR_*`, `Mitsubishi_FX5U_*`, `Mitsubishi_FX3U_ENET_*`,
+  `Mitsubishi_FX3U_ENET_P502_*`. iQ-R built-in Ethernet and the RJ71EN71
+  behave identically for MODBUS/TCP slave purposes and can share a file
+  `Mitsubishi_iQR_*`.
+- **Cross-model shared tests** (word order CDAB, binary not BCD, standard
+  exception codes, 125-register FC03 cap) can live in a single
+  `Mitsubishi_Common_*` fixture.
+
+## References
+
+1. Mitsubishi Electric, *MODBUS Interface Module User's Manual — QJ71MB91*
+   (SH-080578ENG), RS-232/422/485 MODBUS RTU serial module for MELSEC-Q —
+   https://dl.mitsubishielectric.com/dl/fa/document/manual/plc/sh080578eng/sh080578engk.pdf
+2. Inductive Automation, *Ignition Modbus Driver — Mitsubishi Q / iQ-R word
+   order*, documents CDAB convention —
+   https://docs.inductiveautomation.com/docs/8.1/ignition-modules/opc-ua/drivers/modbus-v2
+   and forum discussion https://forum.inductiveautomation.com/t/modbus-tcp-device-word-byte-order/65984
+3. Mitsubishi Electric, *Programmable Controller User's Manual QJ71MB91 MODBUS
+   Interface Module*, Chapter 7 "Parameter Setting" describing the Modbus
+   Device Assignment Parameter block (assignments 1-16, head-device
+   configuration) —
+   https://www.lcautomation.com/dbdocument/29156/QJ71MB91%20Users%20manual.pdf
+4. Mitsubishi Electric, *FX3U-ENET User's Manual* (JY997D18101), Chapter on
+   MODBUS/TCP communication; function code support and connection limits —
+   https://dl.mitsubishielectric.com/dl/fa/document/manual/plc_fx/jy997d18101/jy997d18101h.pdf
+5. Venus Automation, *Mitsubishi FX3U-ENET-P502 Module — Open Port 502 for
+   Modbus TCP/IP* —
+   https://venusautomation.com.au/mitsubishi-fx3u-enet-p502-module-open-port-502-for-modbus-tcp-ip/
+   and FX3U-ENET-ADP user manual (JY997D45801), which confirms the -ADP
+   variant does not support MODBUS —
+   https://dl.mitsubishielectric.com/dl/fa/document/manual/plc_fx/jy997d45801/jy997d45801h.pdf
+6. XML Control / Ubidots integration notes, *FX3G Modbus* — port-502 trap,
+   D-register mapping default, word order reference —
+   https://sites.google.com/site/xmlcontrol/archive/fx3g-modbus
+   and https://ubidots.com/blog/mitsubishi-plc-as-modbus-tcp-server/
+7. FA Support Me, *Modbus TCP on Built-in Ethernet port in iQ-F and iQ-R* —
+   confirms 16-connection limit on iQ-R, 8 on iQ-F, parameter-driven
+   configuration via GX Works3 —
+   https://www.fasupportme.com/portal/en/kb/articles/modbus-tcp-on-build-in-ethernet-port-in-iq-f-and-iq-r-en
+8. Mitsubishi Electric, *MELSEC iQ-R Ethernet User's Manual (Application)*
+   (SH-081259ENG) and *MELSEC iQ-RJ71EN71 User's Manual* Chapter on
+   "Communications Using Modbus/TCP" —
+   https://www.allied-automation.com/wp-content/uploads/2015/02/MITSUBISHI_manual_plc_iq-r_ethernet_users.pdf
+   and https://www.manualslib.com/manual/1533351/Mitsubishi-Electric-Melsec-Iq-Rj71en71.html?page=109
+9. Mitsubishi Electric, *MODBUS/TCP Interface Module User's Manual — QJ71MT91*
+   (SH-080446ENG), exception codes page 248, device assignment parameter
+   pages 116-124, duplicate-assignment-disables-slave note —
+   https://dl.mitsubishielectric.com/dl/fa/document/manual/plc/sh080446eng/sh080446engj.pdf
+10. Mitsubishi Electric, *MELSEC-L Network Features* — LJ71MT91 documented as
+    L-series equivalent of QJ71MT91 with identical MODBUS/TCP behavior —
+    https://us.mitsubishielectric.com/fa/en/products/cnt/programmable-controllers/melsec-l-series/network/features/
+11. Mitsubishi Electric, *MELSEC iQ-F FX5 User's Manual (MODBUS Communication)*
+    (JY997D56101), Chapter 11 "Modbus/TCP Communication Specifications" —
+    function code max-quantity table, frame specification, device assignment
+    defaults —
+    https://dl.mitsubishielectric.com/dl/fa/document/manual/plcf/jy997d56101/jy997d56101h.pdf
+12. MrPLC forum, *FX5U Modbus-TCP Server (Slave)*, firmware ≥ 1.60 enables
+    native server via parameter; earlier firmware required function block —
+    https://mrplc.com/forums/topic/31883-fx5u-modbus-tcp-server-slave/
+    and Industrial Monitor Direct's "FX5U MODBUS TCP Server Workaround"
+    article (reflects older firmware behavior) —
+    https://industrialmonitordirect.com/blogs/knowledgebase/mitsubishi-fx5u-modbus-tcp-server-configuration-workaround
+13. Mitsubishi Electric, *MELSEC iQ-R MODBUS and MODBUS/TCP Reference Manual —
+    RJ71C24 / RJ71C24-R2* (BCN-P5999-1060) — RJ71C24 is serial RTU only,
+    not TCP —
+    https://dl.mitsubishielectric.com/dl/fa/document/manual/plc/bcn-p5999-1060/bcnp59991060b.pdf
+14. HMS Industrial Networks, *eWON and Mitsubishi FX5U PLC* (KB-0264-00) —
+    documents that FX5U X/Y are octal in GX Works3 but hex when viewed as a
+    Q-series PLC through eWON; the project-level hex/octal toggle —
+    https://hmsnetworks.blob.core.windows.net/www/docs/librariesprovider10/downloads-monitored/manuals/knowledge-base/kb-0264-00-en-ewon-and-mitsubishi-fx5u-plc.pdf
+15. Fernhill Software, *Mitsubishi Melsec PLC Data Address* — documents
+    hex-vs-octal device numbering split across MELSEC families —
+    https://www.fernhillsoftware.com/help/drivers/mitsubishi-melsec/data-address-format.html
+16. Inductive Automation support, *Understanding Mitsubishi PLCs* — D registers
+    store signed 16-bit binary, not BCD; DINT combines two consecutive D
+    registers —
+    https://support.inductiveautomation.com/hc/en-us/articles/16517576753165-Understanding-Mitsubishi-PLCs
+17. Mitsubishi Electric, *FXCPU Structured Programming Manual [Device &
+    Common]* (JY997D26001) — FNC 18 BCD and FNC 19 BIN explicit-conversion
+    instructions confirm binary-by-default storage —
+    https://dl.mitsubishielectric.com/dl/fa/document/manual/plc_fx/jy997d26001/jy997d26001l.pdf

src/ZB.MOM.WW.OtOpcUa.Driver.Modbus/MelsecAddress.cs

@@ -0,0 +1,161 @@
+namespace ZB.MOM.WW.OtOpcUa.Driver.Modbus;
+
+/// <summary>
+///     Mitsubishi MELSEC PLC family selector for address-translation helpers. The Q/L/iQ-R
+///     families write bit-device addresses (X, Y) in <b>hexadecimal</b> in GX Works and the
+///     CPU manuals; the FX and iQ-F families write them in <b>octal</b> (same convention as
+///     AutomationDirect DirectLOGIC). Mixing the two up is the #1 MELSEC driver bug source —
+///     an operator typing <c>X20</c> into a Q-series tag config means decimal 32, but the
+///     same string on an FX3U means decimal 16, so the helper must know the family to route
+///     correctly.
+/// </summary>
+public enum MelsecFamily
+{
+    /// <summary>
+    ///     MELSEC-Q / MELSEC-L / MELSEC iQ-R. X and Y device numbers are interpreted as
+    ///     <b>hexadecimal</b>; <c>X20</c> means decimal 32.
+    /// </summary>
+    Q_L_iQR,
+
+    /// <summary>
+    ///     MELSEC-F (FX3U / FX3GE / FX3G) and MELSEC iQ-F (FX5U). X and Y device numbers
+    ///     are interpreted as <b>octal</b> (same as DirectLOGIC); <c>X20</c> means decimal 16.
+    ///     iQ-F has a GX Works3 project toggle that can flip to decimal — if a site uses
+    ///     that, configure the tag's Address directly as a decimal PDU address and do not
+    ///     route through this helper.
+    /// </summary>
+    F_iQF,
+}
+
+/// <summary>
+///     Mitsubishi MELSEC address-translation helpers for the QJ71MT91 / LJ71MT91 / RJ71EN71 /
+///     iQ-R built-in / iQ-F / FX3U-ENET-P502 Modbus modules. MELSEC does NOT hard-wire
+///     Modbus-to-device mappings like DL260 does — every site configures its own "Modbus
+///     Device Assignment Parameter" block of up to 16 entries. The helpers here cover only
+///     the <b>address-notation</b> portion of the translation (hex X20 vs octal X20 + adding
+///     the bank base); the caller is still responsible for knowing the assignment-block
+///     offset for their site.
+/// </summary>
+/// <remarks>
+///     See <c>docs/v2/mitsubishi.md</c> §device-assignment + §X-Y-hex-trap for the full
+///     matrix and primary-source citations.
+/// </remarks>
+public static class MelsecAddress
+{
+    /// <summary>
+    ///     Translate a MELSEC X-input address (e.g. <c>"X0"</c>, <c>"X10"</c>) to a 0-based
+    ///     Modbus discrete-input address, given the PLC family's address notation (hex or
+    ///     octal) and the Modbus Device Assignment block's X-range base.
+    /// </summary>
+    /// <param name="xAddress">MELSEC X address. <c>X</c> prefix optional, case-insensitive.</param>
+    /// <param name="family">The PLC family — determines whether the trailing digits are hex or octal.</param>
+    /// <param name="xBankBase">
+    ///     0-based Modbus DI address the assignment-block has configured X0 to land at.
+    ///     Typical default on QJ71MT91 sample projects: 0. Pass the site-specific value.
+    /// </param>
+    public static ushort XInputToDiscrete(string xAddress, MelsecFamily family, ushort xBankBase = 0) =>
+        AddFamilyOffset(xBankBase, StripPrefix(xAddress, 'X'), family);
+
+    /// <summary>
+    ///     Translate a MELSEC Y-output address to a 0-based Modbus coil address. Same rules
+    ///     as <see cref="XInputToDiscrete"/> for hex/octal parsing.
+    /// </summary>
+    public static ushort YOutputToCoil(string yAddress, MelsecFamily family, ushort yBankBase = 0) =>
+        AddFamilyOffset(yBankBase, StripPrefix(yAddress, 'Y'), family);
+
+    /// <summary>
+    ///     Translate a MELSEC M-relay address (internal relay) to a 0-based Modbus coil
+    ///     address. M-addresses are <b>decimal</b> on every MELSEC family — unlike X/Y which
+    ///     are hex on Q/L/iQ-R. Includes the bank base that the assignment-block configured.
+    /// </summary>
+    public static ushort MRelayToCoil(string mAddress, ushort mBankBase = 0)
+    {
+        var digits = StripPrefix(mAddress, 'M');
+        if (!ushort.TryParse(digits, out var offset))
+            throw new ArgumentException(
+                $"M-relay address '{mAddress}' is not a valid decimal integer", nameof(mAddress));
+        var result = mBankBase + offset;
+        if (result > ushort.MaxValue)
+            throw new OverflowException($"M-relay {mAddress} + base {mBankBase} exceeds 0xFFFF");
+        return (ushort)result;
+    }
+
+    /// <summary>
+    ///     Translate a MELSEC D-register address (data register) to a 0-based Modbus holding
+    ///     register address. D-addresses are <b>decimal</b>. Default assignment convention is
+    ///     D0 → HR 0 (pass <paramref name="dBankBase"/> = 0); sites with shifted layouts pass
+    ///     their configured base.
+    /// </summary>
+    public static ushort DRegisterToHolding(string dAddress, ushort dBankBase = 0)
+    {
+        var digits = StripPrefix(dAddress, 'D');
+        if (!ushort.TryParse(digits, out var offset))
+            throw new ArgumentException(
+                $"D-register address '{dAddress}' is not a valid decimal integer", nameof(dAddress));
+        var result = dBankBase + offset;
+        if (result > ushort.MaxValue)
+            throw new OverflowException($"D-register {dAddress} + base {dBankBase} exceeds 0xFFFF");
+        return (ushort)result;
+    }
+
+    private static string StripPrefix(string address, char expectedPrefix)
+    {
+        if (string.IsNullOrWhiteSpace(address))
+            throw new ArgumentException("Address must not be empty", nameof(address));
+        var s = address.Trim();
+        if (s.Length > 0 && char.ToUpperInvariant(s[0]) == char.ToUpperInvariant(expectedPrefix))
+            s = s.Substring(1);
+        if (s.Length == 0)
+            throw new ArgumentException($"Address '{address}' has no digits after prefix", nameof(address));
+        return s;
+    }
+
+    private static ushort AddFamilyOffset(ushort baseAddr, string digits, MelsecFamily family)
+    {
+        uint offset = family switch
+        {
+            MelsecFamily.Q_L_iQR => ParseHex(digits),
+            MelsecFamily.F_iQF => ParseOctal(digits),
+            _ => throw new ArgumentOutOfRangeException(nameof(family), family, "Unknown MELSEC family"),
+        };
+        var result = baseAddr + offset;
+        if (result > ushort.MaxValue)
+            throw new OverflowException($"Address {baseAddr}+{offset} exceeds 0xFFFF");
+        return (ushort)result;
+    }
+
+    private static uint ParseHex(string digits)
+    {
+        uint result = 0;
+        foreach (var ch in digits)
+        {
+            uint nibble;
+            if (ch >= '0' && ch <= '9') nibble = (uint)(ch - '0');
+            else if (ch >= 'A' && ch <= 'F') nibble = (uint)(ch - 'A' + 10);
+            else if (ch >= 'a' && ch <= 'f') nibble = (uint)(ch - 'a' + 10);
+            else throw new ArgumentException(
+                $"Address contains non-hex digit '{ch}' — Q/L/iQ-R X/Y addresses are hexadecimal",
+                nameof(digits));
+            result = result * 16 + nibble;
+            if (result > ushort.MaxValue)
+                throw new OverflowException($"Hex address exceeds 0xFFFF");
+        }
+        return result;
+    }
+
+    private static uint ParseOctal(string digits)
+    {
+        uint result = 0;
+        foreach (var ch in digits)
+        {
+            if (ch < '0' || ch > '7')
+                throw new ArgumentException(
+                    $"Address contains non-octal digit '{ch}' — FX/iQ-F X/Y addresses are octal (0-7)",
+                    nameof(digits));
+            result = result * 8 + (uint)(ch - '0');
+            if (result > ushort.MaxValue)
+                throw new OverflowException($"Octal address exceeds 0xFFFF");
+        }
+        return result;
+    }
+}

src/ZB.MOM.WW.OtOpcUa.Driver.OpcUaClient/OpcUaClientDriver.cs

@@ -0,0 +1,393 @@
+using Opc.Ua;
+using Opc.Ua.Client;
+using Opc.Ua.Configuration;
+using ZB.MOM.WW.OtOpcUa.Core.Abstractions;
+
+namespace ZB.MOM.WW.OtOpcUa.Driver.OpcUaClient;
+
+/// <summary>
+///     OPC UA Client (gateway) driver. Opens a <see cref="Session"/> against a remote OPC UA
+///     server and re-exposes its address space through the local OtOpcUa server. PR 66 ships
+///     the scaffold: <see cref="IDriver"/> only (connect / close / health). Browse, read,
+///     write, subscribe, and probe land in PRs 67-69.
+/// </summary>
+/// <remarks>
+///     <para>
+///         Builds its own <see cref="ApplicationConfiguration"/> rather than reusing
+///         <c>Client.Shared</c> — Client.Shared is oriented at the interactive CLI; this
+///         driver is an always-on service component with different session-lifetime needs
+///         (keep-alive monitor, session transfer on reconnect, multi-year uptime).
+///     </para>
+///     <para>
+///         <b>Session lifetime</b>: a single <see cref="Session"/> per driver instance.
+///         Subscriptions multiplex onto that session; SDK reconnect handler takes the session
+///         down and brings it back up on remote-server restart — the driver must re-send
+///         subscriptions + TransferSubscriptions on reconnect to avoid dangling
+///         monitored-item handles. That mechanic lands in PR 69.
+///     </para>
+/// </remarks>
+public sealed class OpcUaClientDriver(OpcUaClientDriverOptions options, string driverInstanceId)
+    : IDriver, IReadable, IWritable, IDisposable, IAsyncDisposable
+{
+    // OPC UA StatusCode constants the driver surfaces for local-side faults. Upstream-server
+    // StatusCodes are passed through verbatim per driver-specs.md §8 "cascading quality" —
+    // downstream clients need to distinguish 'remote source down' from 'local driver failure'.
+    private const uint StatusBadNodeIdInvalid = 0x80330000u;
+    private const uint StatusBadInternalError = 0x80020000u;
+    private const uint StatusBadCommunicationError = 0x80050000u;
+
+    private readonly OpcUaClientDriverOptions _options = options;
+    private readonly SemaphoreSlim _gate = new(1, 1);
+
+    /// <summary>Active OPC UA session. Null until <see cref="InitializeAsync"/> returns cleanly.</summary>
+    internal ISession? Session { get; private set; }
+
+    /// <summary>Per-connection gate. PRs 67+ serialize read/write/browse on this.</summary>
+    internal SemaphoreSlim Gate => _gate;
+
+    private DriverHealth _health = new(DriverState.Unknown, null, null);
+    private bool _disposed;
+
+    public string DriverInstanceId => driverInstanceId;
+    public string DriverType => "OpcUaClient";
+
+    public async Task InitializeAsync(string driverConfigJson, CancellationToken cancellationToken)
+    {
+        _health = new DriverHealth(DriverState.Initializing, null, null);
+        try
+        {
+            var appConfig = await BuildApplicationConfigurationAsync(cancellationToken).ConfigureAwait(false);
+
+            // Endpoint selection: let the stack pick the best matching endpoint for the
+            // requested security policy/mode so the driver doesn't have to hand-validate.
+            // UseSecurity=false when SecurityMode=None shortcuts around cert validation
+            // entirely and is the typical dev-bench configuration.
+            var useSecurity = _options.SecurityMode != OpcUaSecurityMode.None;
+            // The non-obsolete SelectEndpointAsync overloads all require an ITelemetryContext
+            // parameter. Passing null is valid — the SDK falls through to its built-in default
+            // trace sink. Plumbing a telemetry context through every driver surface is out of
+            // scope; the driver emits its own logs via the health surface anyway.
+            var selected = await CoreClientUtils.SelectEndpointAsync(
+                appConfig, _options.EndpointUrl, useSecurity,
+                telemetry: null!,
+                ct: cancellationToken).ConfigureAwait(false);
+            var endpointConfig = EndpointConfiguration.Create(appConfig);
+            endpointConfig.OperationTimeout = (int)_options.Timeout.TotalMilliseconds;
+            var endpoint = new ConfiguredEndpoint(null, selected, endpointConfig);
+
+            var identity = _options.AuthType switch
+            {
+                OpcUaAuthType.Anonymous => new UserIdentity(new AnonymousIdentityToken()),
+                // The UserIdentity(string, string) overload was removed in favour of
+                // (string, byte[]) to make the password encoding explicit. UTF-8 is the
+                // overwhelmingly common choice for Basic256Sha256-secured sessions.
+                OpcUaAuthType.Username => new UserIdentity(
+                    _options.Username ?? string.Empty,
+                    System.Text.Encoding.UTF8.GetBytes(_options.Password ?? string.Empty)),
+                OpcUaAuthType.Certificate => throw new NotSupportedException(
+                    "Certificate authentication lands in a follow-up PR; for now use Anonymous or Username"),
+                _ => new UserIdentity(new AnonymousIdentityToken()),
+            };
+
+            // All Session.Create* static methods are marked [Obsolete] in SDK 1.5.378; the
+            // non-obsolete path is DefaultSessionFactory.Instance.CreateAsync (which is the
+            // 8-arg signature matching our driver config — ApplicationConfiguration +
+            // ConfiguredEndpoint, no transport-waiting-connection or reverse-connect-manager
+            // required for the standard opc.tcp direct-connect case).
+            // DefaultSessionFactory's parameterless ctor is also obsolete in 1.5.378; the
+            // current constructor requires an ITelemetryContext. Passing null is tolerated —
+            // the factory falls back to its internal default sink, same as the telemetry:null
+            // on SelectEndpointAsync above.
+            var session = await new DefaultSessionFactory(telemetry: null!).CreateAsync(
+                appConfig,
+                endpoint,
+                false,                                              // updateBeforeConnect
+                _options.SessionName,
+                (uint)_options.SessionTimeout.TotalMilliseconds,
+                identity,
+                null,                                               // preferredLocales
+                cancellationToken).ConfigureAwait(false);
+
+            session.KeepAliveInterval = (int)_options.KeepAliveInterval.TotalMilliseconds;
+
+            Session = session;
+            _health = new DriverHealth(DriverState.Healthy, DateTime.UtcNow, null);
+        }
+        catch (Exception ex)
+        {
+            try { if (Session is Session s) await s.CloseAsync().ConfigureAwait(false); } catch { }
+            Session = null;
+            _health = new DriverHealth(DriverState.Faulted, null, ex.Message);
+            throw;
+        }
+    }
+
+    /// <summary>
+    ///     Build a minimal in-memory <see cref="ApplicationConfiguration"/>. Certificates live
+    ///     under the OS user profile — on Windows that's <c>%LocalAppData%\OtOpcUa\pki</c>
+    ///     — so multiple driver instances in the same OtOpcUa server process share one
+    ///     certificate store without extra config.
+    /// </summary>
+    private async Task<ApplicationConfiguration> BuildApplicationConfigurationAsync(CancellationToken ct)
+    {
+        // The default ctor is obsolete in favour of the ITelemetryContext overload; suppress
+        // locally rather than plumbing a telemetry context all the way through the driver
+        // surface — the driver emits no per-request telemetry of its own and the SDK's
+        // internal fallback is fine for a gateway use case.
+#pragma warning disable CS0618
+        var app = new ApplicationInstance
+        {
+            ApplicationName = _options.SessionName,
+            ApplicationType = ApplicationType.Client,
+        };
+#pragma warning restore CS0618
+
+        var pkiRoot = Path.Combine(
+            Environment.GetFolderPath(Environment.SpecialFolder.LocalApplicationData),
+            "OtOpcUa", "pki");
+
+        var config = new ApplicationConfiguration
+        {
+            ApplicationName = _options.SessionName,
+            ApplicationType = ApplicationType.Client,
+            ApplicationUri = _options.ApplicationUri,
+            SecurityConfiguration = new SecurityConfiguration
+            {
+                ApplicationCertificate = new CertificateIdentifier
+                {
+                    StoreType = CertificateStoreType.Directory,
+                    StorePath = Path.Combine(pkiRoot, "own"),
+                    SubjectName = $"CN={_options.SessionName}",
+                },
+                TrustedPeerCertificates = new CertificateTrustList
+                {
+                    StoreType = CertificateStoreType.Directory,
+                    StorePath = Path.Combine(pkiRoot, "trusted"),
+                },
+                TrustedIssuerCertificates = new CertificateTrustList
+                {
+                    StoreType = CertificateStoreType.Directory,
+                    StorePath = Path.Combine(pkiRoot, "issuers"),
+                },
+                RejectedCertificateStore = new CertificateTrustList
+                {
+                    StoreType = CertificateStoreType.Directory,
+                    StorePath = Path.Combine(pkiRoot, "rejected"),
+                },
+                AutoAcceptUntrustedCertificates = _options.AutoAcceptCertificates,
+            },
+            TransportQuotas = new TransportQuotas { OperationTimeout = (int)_options.Timeout.TotalMilliseconds },
+            ClientConfiguration = new ClientConfiguration
+            {
+                DefaultSessionTimeout = (int)_options.SessionTimeout.TotalMilliseconds,
+            },
+            DisableHiResClock = true,
+        };
+
+        await config.ValidateAsync(ApplicationType.Client, ct).ConfigureAwait(false);
+
+        // Attach a cert-validator handler that honours the AutoAccept flag. Without this,
+        // AutoAcceptUntrustedCertificates on the config alone isn't always enough in newer
+        // SDK versions — the validator raises an event the app has to handle.
+        if (_options.AutoAcceptCertificates)
+        {
+            config.CertificateValidator.CertificateValidation += (s, e) =>
+            {
+                if (e.Error.StatusCode == StatusCodes.BadCertificateUntrusted)
+                    e.Accept = true;
+            };
+        }
+
+        // Ensure an application certificate exists. The SDK auto-generates one if missing.
+        app.ApplicationConfiguration = config;
+        await app.CheckApplicationInstanceCertificatesAsync(silent: true, lifeTimeInMonths: null, ct)
+            .ConfigureAwait(false);
+
+        return config;
+    }
+
+    public async Task ReinitializeAsync(string driverConfigJson, CancellationToken cancellationToken)
+    {
+        await ShutdownAsync(cancellationToken).ConfigureAwait(false);
+        await InitializeAsync(driverConfigJson, cancellationToken).ConfigureAwait(false);
+    }
+
+    public async Task ShutdownAsync(CancellationToken cancellationToken)
+    {
+        try { if (Session is Session s) await s.CloseAsync(cancellationToken).ConfigureAwait(false); }
+        catch { /* best-effort */ }
+        try { Session?.Dispose(); } catch { }
+        Session = null;
+        _health = new DriverHealth(DriverState.Unknown, _health.LastSuccessfulRead, null);
+    }
+
+    public DriverHealth GetHealth() => _health;
+    public long GetMemoryFootprint() => 0;
+    public Task FlushOptionalCachesAsync(CancellationToken cancellationToken) => Task.CompletedTask;
+
+    // ---- IReadable ----
+
+    public async Task<IReadOnlyList<DataValueSnapshot>> ReadAsync(
+        IReadOnlyList<string> fullReferences, CancellationToken cancellationToken)
+    {
+        var session = RequireSession();
+        var results = new DataValueSnapshot[fullReferences.Count];
+        var now = DateTime.UtcNow;
+
+        // Parse NodeIds up-front. Tags whose reference doesn't parse get BadNodeIdInvalid
+        // and are omitted from the wire request — saves a round-trip against the upstream
+        // server for a fault the driver can detect locally.
+        var toSend = new ReadValueIdCollection();
+        var indexMap = new List<int>(fullReferences.Count); // maps wire-index -> results-index
+        for (var i = 0; i < fullReferences.Count; i++)
+        {
+            if (!TryParseNodeId(session, fullReferences[i], out var nodeId))
+            {
+                results[i] = new DataValueSnapshot(null, StatusBadNodeIdInvalid, null, now);
+                continue;
+            }
+            toSend.Add(new ReadValueId { NodeId = nodeId, AttributeId = Attributes.Value });
+            indexMap.Add(i);
+        }
+
+        if (toSend.Count == 0) return results;
+
+        await _gate.WaitAsync(cancellationToken).ConfigureAwait(false);
+        try
+        {
+            try
+            {
+                var resp = await session.ReadAsync(
+                    requestHeader: null,
+                    maxAge: 0,
+                    timestampsToReturn: TimestampsToReturn.Both,
+                    nodesToRead: toSend,
+                    ct: cancellationToken).ConfigureAwait(false);
+
+                var values = resp.Results;
+                for (var w = 0; w < values.Count; w++)
+                {
+                    var r = indexMap[w];
+                    var dv = values[w];
+                    // Preserve the upstream StatusCode verbatim — including Bad codes per
+                    // §8's cascading-quality rule. Also preserve SourceTimestamp so downstream
+                    // clients can detect stale upstream data.
+                    results[r] = new DataValueSnapshot(
+                        Value: dv.Value,
+                        StatusCode: dv.StatusCode.Code,
+                        SourceTimestampUtc: dv.SourceTimestamp == DateTime.MinValue ? null : dv.SourceTimestamp,
+                        ServerTimestampUtc: dv.ServerTimestamp == DateTime.MinValue ? now : dv.ServerTimestamp);
+                }
+                _health = new DriverHealth(DriverState.Healthy, now, null);
+            }
+            catch (Exception ex)
+            {
+                // Transport / timeout / session-dropped — fan out the same fault across every
+                // tag in this batch. Per-tag StatusCode stays BadCommunicationError (not
+                // BadInternalError) so operators distinguish "upstream unreachable" from
+                // "driver bug".
+                for (var w = 0; w < indexMap.Count; w++)
+                {
+                    var r = indexMap[w];
+                    results[r] = new DataValueSnapshot(null, StatusBadCommunicationError, null, now);
+                }
+                _health = new DriverHealth(DriverState.Degraded, _health.LastSuccessfulRead, ex.Message);
+            }
+        }
+        finally { _gate.Release(); }
+        return results;
+    }
+
+    // ---- IWritable ----
+
+    public async Task<IReadOnlyList<WriteResult>> WriteAsync(
+        IReadOnlyList<Core.Abstractions.WriteRequest> writes, CancellationToken cancellationToken)
+    {
+        var session = RequireSession();
+        var results = new WriteResult[writes.Count];
+
+        var toSend = new WriteValueCollection();
+        var indexMap = new List<int>(writes.Count);
+        for (var i = 0; i < writes.Count; i++)
+        {
+            if (!TryParseNodeId(session, writes[i].FullReference, out var nodeId))
+            {
+                results[i] = new WriteResult(StatusBadNodeIdInvalid);
+                continue;
+            }
+            toSend.Add(new WriteValue
+            {
+                NodeId = nodeId,
+                AttributeId = Attributes.Value,
+                Value = new DataValue(new Variant(writes[i].Value)),
+            });
+            indexMap.Add(i);
+        }
+
+        if (toSend.Count == 0) return results;
+
+        await _gate.WaitAsync(cancellationToken).ConfigureAwait(false);
+        try
+        {
+            try
+            {
+                var resp = await session.WriteAsync(
+                    requestHeader: null,
+                    nodesToWrite: toSend,
+                    ct: cancellationToken).ConfigureAwait(false);
+
+                var codes = resp.Results;
+                for (var w = 0; w < codes.Count; w++)
+                {
+                    var r = indexMap[w];
+                    // Pass upstream WriteResult StatusCode through verbatim. Success codes
+                    // include Good (0) and any warning-level Good* variants; anything with
+                    // the severity bits set is a Bad.
+                    results[r] = new WriteResult(codes[w].Code);
+                }
+            }
+            catch (Exception)
+            {
+                for (var w = 0; w < indexMap.Count; w++)
+                    results[indexMap[w]] = new WriteResult(StatusBadCommunicationError);
+            }
+        }
+        finally { _gate.Release(); }
+        return results;
+    }
+
+    /// <summary>
+    ///     Parse a tag's full-reference string as a NodeId. Accepts the standard OPC UA
+    ///     serialized forms (<c>ns=2;s=…</c>, <c>i=2253</c>, <c>ns=4;g=…</c>, <c>ns=3;b=…</c>).
+    ///     Empty + malformed strings return false; the driver surfaces that as
+    ///     <see cref="StatusBadNodeIdInvalid"/> without a wire round-trip.
+    /// </summary>
+    internal static bool TryParseNodeId(ISession session, string fullReference, out NodeId nodeId)
+    {
+        nodeId = NodeId.Null;
+        if (string.IsNullOrWhiteSpace(fullReference)) return false;
+        try
+        {
+            nodeId = NodeId.Parse(session.MessageContext, fullReference);
+            return !NodeId.IsNull(nodeId);
+        }
+        catch
+        {
+            return false;
+        }
+    }
+
+    private ISession RequireSession() =>
+        Session ?? throw new InvalidOperationException("OpcUaClientDriver not initialized");
+
+    public void Dispose() => DisposeAsync().AsTask().GetAwaiter().GetResult();
+
+    public async ValueTask DisposeAsync()
+    {
+        if (_disposed) return;
+        _disposed = true;
+        try { await ShutdownAsync(CancellationToken.None).ConfigureAwait(false); }
+        catch { /* disposal is best-effort */ }
+        _gate.Dispose();
+    }
+}

src/ZB.MOM.WW.OtOpcUa.Driver.OpcUaClient/OpcUaClientDriverOptions.cs

@@ -0,0 +1,81 @@
+namespace ZB.MOM.WW.OtOpcUa.Driver.OpcUaClient;
+
+/// <summary>
+///     OPC UA Client (gateway) driver configuration. Bound from <c>DriverConfig</c> JSON at
+///     driver-host registration time. Models the settings documented in
+///     <c>docs/v2/driver-specs.md</c> §8.
+/// </summary>
+/// <remarks>
+///     This driver connects to a REMOTE OPC UA server and re-exposes its address space
+///     through the local OtOpcUa server — the opposite direction from the usual "server
+///     exposes PLC data" flow. Tier A (pure managed, OPC Foundation reference SDK); universal
+///     protections cover it.
+/// </remarks>
+public sealed class OpcUaClientDriverOptions
+{
+    /// <summary>Remote OPC UA endpoint URL, e.g. <c>opc.tcp://plc.internal:4840</c>.</summary>
+    public string EndpointUrl { get; init; } = "opc.tcp://localhost:4840";
+
+    /// <summary>Security policy. One of <c>None</c>, <c>Basic256Sha256</c>, <c>Aes128_Sha256_RsaOaep</c>.</summary>
+    public string SecurityPolicy { get; init; } = "None";
+
+    /// <summary>Security mode.</summary>
+    public OpcUaSecurityMode SecurityMode { get; init; } = OpcUaSecurityMode.None;
+
+    /// <summary>Authentication type.</summary>
+    public OpcUaAuthType AuthType { get; init; } = OpcUaAuthType.Anonymous;
+
+    /// <summary>User name (required only for <see cref="OpcUaAuthType.Username"/>).</summary>
+    public string? Username { get; init; }
+
+    /// <summary>Password (required only for <see cref="OpcUaAuthType.Username"/>).</summary>
+    public string? Password { get; init; }
+
+    /// <summary>Server-negotiated session timeout. Default 120s per driver-specs.md §8.</summary>
+    public TimeSpan SessionTimeout { get; init; } = TimeSpan.FromSeconds(120);
+
+    /// <summary>Client-side keep-alive interval.</summary>
+    public TimeSpan KeepAliveInterval { get; init; } = TimeSpan.FromSeconds(5);
+
+    /// <summary>Initial reconnect delay after a session drop.</summary>
+    public TimeSpan ReconnectPeriod { get; init; } = TimeSpan.FromSeconds(5);
+
+    /// <summary>
+    ///     When <c>true</c>, the driver accepts any self-signed / untrusted server certificate.
+    ///     Dev-only — must be <c>false</c> in production so MITM attacks against the opc.tcp
+    ///     channel fail closed.
+    /// </summary>
+    public bool AutoAcceptCertificates { get; init; } = false;
+
+    /// <summary>
+    ///     Application URI the driver reports during session creation. Must match the
+    ///     subject-alt-name on the client certificate if one is used, which is why it's a
+    ///     config knob rather than hard-coded.
+    /// </summary>
+    public string ApplicationUri { get; init; } = "urn:localhost:OtOpcUa:GatewayClient";
+
+    /// <summary>
+    ///     Friendly name sent to the remote server for diagnostics. Shows up in the remote
+    ///     server's session-list so operators can identify which gateway instance is calling.
+    /// </summary>
+    public string SessionName { get; init; } = "OtOpcUa-Gateway";
+
+    /// <summary>Connect + per-operation timeout.</summary>
+    public TimeSpan Timeout { get; init; } = TimeSpan.FromSeconds(10);
+}
+
+/// <summary>OPC UA message security mode.</summary>
+public enum OpcUaSecurityMode
+{
+    None,
+    Sign,
+    SignAndEncrypt,
+}
+
+/// <summary>User authentication type sent to the remote server.</summary>
+public enum OpcUaAuthType
+{
+    Anonymous,
+    Username,
+    Certificate,
+}

src/ZB.MOM.WW.OtOpcUa.Driver.OpcUaClient/ZB.MOM.WW.OtOpcUa.Driver.OpcUaClient.csproj

@@ -0,0 +1,28 @@
+<Project Sdk="Microsoft.NET.Sdk">
+
+    <PropertyGroup>
+        <TargetFramework>net10.0</TargetFramework>
+        <Nullable>enable</Nullable>
+        <ImplicitUsings>enable</ImplicitUsings>
+        <LangVersion>latest</LangVersion>
+        <TreatWarningsAsErrors>true</TreatWarningsAsErrors>
+        <GenerateDocumentationFile>true</GenerateDocumentationFile>
+        <NoWarn>$(NoWarn);CS1591</NoWarn>
+        <RootNamespace>ZB.MOM.WW.OtOpcUa.Driver.OpcUaClient</RootNamespace>
+        <AssemblyName>ZB.MOM.WW.OtOpcUa.Driver.OpcUaClient</AssemblyName>
+    </PropertyGroup>
+
+    <ItemGroup>
+        <ProjectReference Include="..\ZB.MOM.WW.OtOpcUa.Core.Abstractions\ZB.MOM.WW.OtOpcUa.Core.Abstractions.csproj"/>
+    </ItemGroup>
+
+    <ItemGroup>
+        <PackageReference Include="OPCFoundation.NetStandard.Opc.Ua.Client" Version="1.5.378.106"/>
+        <PackageReference Include="OPCFoundation.NetStandard.Opc.Ua.Configuration" Version="1.5.378.106"/>
+    </ItemGroup>
+
+    <ItemGroup>
+        <InternalsVisibleTo Include="ZB.MOM.WW.OtOpcUa.Driver.OpcUaClient.Tests"/>
+    </ItemGroup>
+
+</Project>

src/ZB.MOM.WW.OtOpcUa.Driver.S7/S7AddressParser.cs

@@ -0,0 +1,216 @@
+namespace ZB.MOM.WW.OtOpcUa.Driver.S7;
+
+/// <summary>
+///     Siemens S7 memory area. The driver's tag-address parser maps every S7 tag string into
+///     exactly one of these + an offset. Values match the on-wire S7 area codes only
+///     incidentally — S7.Net uses its own <c>DataType</c> enum (<c>DataBlock</c>, <c>Memory</c>,
+///     <c>Input</c>, <c>Output</c>, <c>Timer</c>, <c>Counter</c>) so the adapter layer translates.
+/// </summary>
+public enum S7Area
+{
+    DataBlock,
+    Memory,  // M (Merker / marker byte)
+    Input,   // I (process-image input)
+    Output,  // Q (process-image output)
+    Timer,
+    Counter,
+}
+
+/// <summary>
+///     Access width for a DB / M / I / Q address. Timers and counters are always 16-bit
+///     opaque (not user-addressable via size suffixes).
+/// </summary>
+public enum S7Size
+{
+    Bit,     // X
+    Byte,    // B
+    Word,    // W — 16-bit
+    DWord,   // D — 32-bit
+}
+
+/// <summary>
+///     Parsed form of an S7 tag-address string. Produced by <see cref="S7AddressParser.Parse"/>.
+/// </summary>
+/// <param name="Area">Memory area (DB, M, I, Q, T, C).</param>
+/// <param name="DbNumber">Data block number; only meaningful when <paramref name="Area"/> is <see cref="S7Area.DataBlock"/>.</param>
+/// <param name="Size">Access width. Always <see cref="S7Size.Word"/> for Timer and Counter.</param>
+/// <param name="ByteOffset">Byte offset into the area (for DB/M/I/Q) or the timer/counter number.</param>
+/// <param name="BitOffset">Bit position 0-7 when <paramref name="Size"/> is <see cref="S7Size.Bit"/>; 0 otherwise.</param>
+public readonly record struct S7ParsedAddress(
+    S7Area Area,
+    int DbNumber,
+    S7Size Size,
+    int ByteOffset,
+    int BitOffset);
+
+/// <summary>
+///     Parses Siemens S7 address strings into <see cref="S7ParsedAddress"/>. Accepts the
+///     Siemens TIA-Portal / STEP 7 Classic syntax documented in <c>docs/v2/driver-specs.md</c> §5:
+///     <list type="bullet">
+///         <item><c>DB{n}.DB{X|B|W|D}{offset}[.bit]</c> — e.g. <c>DB1.DBX0.0</c>, <c>DB1.DBW0</c>, <c>DB1.DBD4</c></item>
+///         <item><c>M{B|W|D}{offset}</c> or <c>M{offset}.{bit}</c> — e.g. <c>MB0</c>, <c>MW0</c>, <c>MD4</c>, <c>M0.0</c></item>
+///         <item><c>I{B|W|D}{offset}</c> or <c>I{offset}.{bit}</c> — e.g. <c>IB0</c>, <c>IW0</c>, <c>ID0</c>, <c>I0.0</c></item>
+///         <item><c>Q{B|W|D}{offset}</c> or <c>Q{offset}.{bit}</c> — e.g. <c>QB0</c>, <c>QW0</c>, <c>QD0</c>, <c>Q0.0</c></item>
+///         <item><c>T{n}</c> — e.g. <c>T0</c>, <c>T15</c></item>
+///         <item><c>C{n}</c> — e.g. <c>C0</c>, <c>C10</c></item>
+///     </list>
+///     Grammar is case-insensitive. Leading/trailing whitespace tolerated. Bit specifiers
+///     must be 0-7; byte offsets must be non-negative; DB numbers must be &gt;= 1.
+/// </summary>
+/// <remarks>
+///     Parse is deliberately strict — the parser rejects syntactic garbage up-front so a bad
+///     tag config fails at driver init time instead of surfacing as a misleading
+///     <c>BadInternalError</c> on every Read against that tag.
+/// </remarks>
+public static class S7AddressParser
+{
+    /// <summary>
+    ///     Parse an S7 address. Throws <see cref="FormatException"/> on any syntax error with
+    ///     the offending input echoed in the message so operators can correlate to the tag
+    ///     config that produced the fault.
+    /// </summary>
+    public static S7ParsedAddress Parse(string address)
+    {
+        if (string.IsNullOrWhiteSpace(address))
+            throw new FormatException("S7 address must not be empty");
+        var s = address.Trim().ToUpperInvariant();
+
+        // --- DB{n}.DB{X|B|W|D}{offset}[.bit] ---
+        if (s.StartsWith("DB") && TryParseDataBlock(s, out var dbResult))
+            return dbResult;
+
+        if (s.Length < 2)
+            throw new FormatException($"S7 address '{address}' is too short to parse");
+
+        var areaChar = s[0];
+        var rest = s.Substring(1);
+
+        switch (areaChar)
+        {
+            case 'M': return ParseMIQ(S7Area.Memory, rest, address);
+            case 'I': return ParseMIQ(S7Area.Input, rest, address);
+            case 'Q': return ParseMIQ(S7Area.Output, rest, address);
+            case 'T': return ParseTimerOrCounter(S7Area.Timer, rest, address);
+            case 'C': return ParseTimerOrCounter(S7Area.Counter, rest, address);
+            default:
+                throw new FormatException($"S7 address '{address}' starts with unknown area '{areaChar}' (expected DB/M/I/Q/T/C)");
+        }
+    }
+
+    /// <summary>
+    ///     Try-parse variant for callers that can't afford an exception on bad input (e.g.
+    ///     config validation pages in the Admin UI). Returns <c>false</c> for any input that
+    ///     would throw from <see cref="Parse"/>.
+    /// </summary>
+    public static bool TryParse(string address, out S7ParsedAddress result)
+    {
+        try
+        {
+            result = Parse(address);
+            return true;
+        }
+        catch (FormatException)
+        {
+            result = default;
+            return false;
+        }
+    }
+
+    private static bool TryParseDataBlock(string s, out S7ParsedAddress result)
+    {
+        result = default;
+        // Split on first '.': left side must be DB{n}, right side DB{X|B|W|D}{offset}[.bit]
+        var dot = s.IndexOf('.');
+        if (dot < 0) return false;
+        var head = s.Substring(0, dot);    // DB{n}
+        var tail = s.Substring(dot + 1);   // DB{X|B|W|D}{offset}[.bit]
+
+        if (head.Length < 3) return false;
+        if (!int.TryParse(head.AsSpan(2), out var dbNumber) || dbNumber < 1)
+            throw new FormatException($"S7 DB number in '{s}' must be a positive integer");
+
+        if (!tail.StartsWith("DB") || tail.Length < 4)
+            throw new FormatException($"S7 DB address tail '{tail}' must start with DB{{X|B|W|D}}");
+
+        var sizeChar = tail[2];
+        var offsetStart = 3;
+        var size = sizeChar switch
+        {
+            'X' => S7Size.Bit,
+            'B' => S7Size.Byte,
+            'W' => S7Size.Word,
+            'D' => S7Size.DWord,
+            _ => throw new FormatException($"S7 DB size '{sizeChar}' in '{s}' must be X/B/W/D"),
+        };
+
+        var (byteOffset, bitOffset) = ParseOffsetAndOptionalBit(tail, offsetStart, size, s);
+        result = new S7ParsedAddress(S7Area.DataBlock, dbNumber, size, byteOffset, bitOffset);
+        return true;
+    }
+
+    private static S7ParsedAddress ParseMIQ(S7Area area, string rest, string original)
+    {
+        if (rest.Length == 0)
+            throw new FormatException($"S7 address '{original}' has no offset");
+
+        var first = rest[0];
+        S7Size size;
+        int offsetStart;
+        switch (first)
+        {
+            case 'B': size = S7Size.Byte;  offsetStart = 1; break;
+            case 'W': size = S7Size.Word;  offsetStart = 1; break;
+            case 'D': size = S7Size.DWord; offsetStart = 1; break;
+            default:
+                // No size prefix => bit-level address requires explicit .bit. Size stays Bit;
+                // ParseOffsetAndOptionalBit will demand the dot.
+                size = S7Size.Bit;
+                offsetStart = 0;
+                break;
+        }
+
+        var (byteOffset, bitOffset) = ParseOffsetAndOptionalBit(rest, offsetStart, size, original);
+        return new S7ParsedAddress(area, DbNumber: 0, size, byteOffset, bitOffset);
+    }
+
+    private static S7ParsedAddress ParseTimerOrCounter(S7Area area, string rest, string original)
+    {
+        if (rest.Length == 0)
+            throw new FormatException($"S7 address '{original}' has no {area} number");
+        if (!int.TryParse(rest, out var number) || number < 0)
+            throw new FormatException($"S7 {area} number in '{original}' must be a non-negative integer");
+        return new S7ParsedAddress(area, DbNumber: 0, S7Size.Word, number, BitOffset: 0);
+    }
+
+    private static (int byteOffset, int bitOffset) ParseOffsetAndOptionalBit(
+        string s, int start, S7Size size, string original)
+    {
+        var offsetEnd = start;
+        while (offsetEnd < s.Length && s[offsetEnd] >= '0' && s[offsetEnd] <= '9')
+            offsetEnd++;
+        if (offsetEnd == start)
+            throw new FormatException($"S7 address '{original}' has no byte-offset digits");
+
+        if (!int.TryParse(s.AsSpan(start, offsetEnd - start), out var byteOffset) || byteOffset < 0)
+            throw new FormatException($"S7 byte offset in '{original}' must be non-negative");
+
+        // No bit-suffix: done unless size is Bit with no prefix, which requires one.
+        if (offsetEnd == s.Length)
+        {
+            if (size == S7Size.Bit)
+                throw new FormatException($"S7 address '{original}' needs a .{{bit}} suffix for bit access");
+            return (byteOffset, 0);
+        }
+
+        if (s[offsetEnd] != '.')
+            throw new FormatException($"S7 address '{original}' has unexpected character after offset");
+
+        if (size != S7Size.Bit)
+            throw new FormatException($"S7 address '{original}' has a bit suffix but the size is {size} — bit access needs X (DB) or no size prefix (M/I/Q)");
+
+        if (!int.TryParse(s.AsSpan(offsetEnd + 1), out var bitOffset) || bitOffset is < 0 or > 7)
+            throw new FormatException($"S7 bit offset in '{original}' must be 0-7");
+
+        return (byteOffset, bitOffset);
+    }
+}

src/ZB.MOM.WW.OtOpcUa.Driver.S7/S7Driver.cs

@@ -0,0 +1,513 @@
+using S7.Net;
+using ZB.MOM.WW.OtOpcUa.Core.Abstractions;
+
+namespace ZB.MOM.WW.OtOpcUa.Driver.S7;
+
+/// <summary>
+///     Siemens S7 native driver — speaks S7comm over ISO-on-TCP (port 102) via the S7netplus
+///     library. First implementation of <see cref="IDriver"/> for an in-process .NET Standard
+///     PLC protocol that is NOT Modbus, validating that the v2 driver-capability interfaces
+///     generalize beyond Modbus + Galaxy.
+/// </summary>
+/// <remarks>
+///     <para>
+///         PR 62 ships the scaffold: <see cref="IDriver"/> only (Initialize / Reinitialize /
+///         Shutdown / GetHealth). <see cref="ITagDiscovery"/>, <see cref="IReadable"/>,
+///         <see cref="IWritable"/>, <see cref="ISubscribable"/>, <see cref="IHostConnectivityProbe"/>
+///         land in PRs 63-65 once the address parser (PR 63) is in place.
+///     </para>
+///     <para>
+///         <b>Single-connection policy</b>: S7netplus documented pattern is one
+///         <c>Plc</c> instance per PLC, serialized with a <see cref="SemaphoreSlim"/>.
+///         Parallelising reads against a single S7 CPU doesn't help — the CPU scans the
+///         communication mailbox at most once per cycle (2-10 ms) and queues concurrent
+///         requests wire-side anyway. Multiple client-side connections just waste the CPU's
+///         8-64 connection-resource budget.
+///     </para>
+/// </remarks>
+public sealed class S7Driver(S7DriverOptions options, string driverInstanceId)
+    : IDriver, ITagDiscovery, IReadable, IWritable, ISubscribable, IHostConnectivityProbe, IDisposable, IAsyncDisposable
+{
+    // ---- ISubscribable + IHostConnectivityProbe state ----
+
+    private readonly System.Collections.Concurrent.ConcurrentDictionary<long, SubscriptionState> _subscriptions = new();
+    private long _nextSubscriptionId;
+    private readonly object _probeLock = new();
+    private HostState _hostState = HostState.Unknown;
+    private DateTime _hostStateChangedUtc = DateTime.UtcNow;
+    private CancellationTokenSource? _probeCts;
+
+    public event EventHandler<DataChangeEventArgs>? OnDataChange;
+    public event EventHandler<HostStatusChangedEventArgs>? OnHostStatusChanged;
+
+    /// <summary>OPC UA StatusCode used when the tag name isn't in the driver's tag map.</summary>
+    private const uint StatusBadNodeIdUnknown = 0x80340000u;
+    /// <summary>OPC UA StatusCode used when the tag's data type isn't implemented yet.</summary>
+    private const uint StatusBadNotSupported = 0x803D0000u;
+    /// <summary>OPC UA StatusCode used when the tag is declared read-only.</summary>
+    private const uint StatusBadNotWritable = 0x803B0000u;
+    /// <summary>OPC UA StatusCode used when write fails validation (e.g. out-of-range value).</summary>
+    private const uint StatusBadInternalError = 0x80020000u;
+    /// <summary>OPC UA StatusCode used for socket / timeout / protocol-layer faults.</summary>
+    private const uint StatusBadCommunicationError = 0x80050000u;
+    /// <summary>OPC UA StatusCode used when S7 returns <c>ErrorCode.WrongCPU</c> / PUT/GET disabled.</summary>
+    private const uint StatusBadDeviceFailure = 0x80550000u;
+
+    private readonly Dictionary<string, S7TagDefinition> _tagsByName = new(StringComparer.OrdinalIgnoreCase);
+    private readonly Dictionary<string, S7ParsedAddress> _parsedByName = new(StringComparer.OrdinalIgnoreCase);
+
+    private readonly S7DriverOptions _options = options;
+    private readonly SemaphoreSlim _gate = new(1, 1);
+
+    /// <summary>
+    ///     Per-connection gate. Internal so PRs 63-65 (read/write/subscribe) can serialize on
+    ///     the same semaphore without exposing it publicly. Single-connection-per-PLC is a
+    ///     hard requirement of S7netplus — see class remarks.
+    /// </summary>
+    internal SemaphoreSlim Gate => _gate;
+
+    /// <summary>
+    ///     Active S7.Net PLC connection. Null until <see cref="InitializeAsync"/> returns; null
+    ///     after <see cref="ShutdownAsync"/>. Read-only outside this class; PR 64's Read/Write
+    ///     will take the <see cref="_gate"/> before touching it.
+    /// </summary>
+    internal Plc? Plc { get; private set; }
+
+    private DriverHealth _health = new(DriverState.Unknown, null, null);
+    private bool _disposed;
+
+    public string DriverInstanceId => driverInstanceId;
+    public string DriverType => "S7";
+
+    public async Task InitializeAsync(string driverConfigJson, CancellationToken cancellationToken)
+    {
+        _health = new DriverHealth(DriverState.Initializing, null, null);
+        try
+        {
+            var plc = new Plc(_options.CpuType, _options.Host, _options.Rack, _options.Slot);
+            // S7netplus writes timeouts into the underlying TcpClient via Plc.WriteTimeout /
+            // Plc.ReadTimeout (milliseconds). Set before OpenAsync so the handshake itself
+            // honours the bound.
+            plc.WriteTimeout = (int)_options.Timeout.TotalMilliseconds;
+            plc.ReadTimeout = (int)_options.Timeout.TotalMilliseconds;
+
+            using var cts = CancellationTokenSource.CreateLinkedTokenSource(cancellationToken);
+            cts.CancelAfter(_options.Timeout);
+            await plc.OpenAsync(cts.Token).ConfigureAwait(false);
+
+            Plc = plc;
+
+            // Parse every tag's address once at init so config typos fail fast here instead
+            // of surfacing as BadInternalError on every Read against the bad tag. The parser
+            // also rejects bit-offset > 7, DB 0, unknown area letters, etc.
+            _tagsByName.Clear();
+            _parsedByName.Clear();
+            foreach (var t in _options.Tags)
+            {
+                var parsed = S7AddressParser.Parse(t.Address); // throws FormatException
+                _tagsByName[t.Name] = t;
+                _parsedByName[t.Name] = parsed;
+            }
+
+            _health = new DriverHealth(DriverState.Healthy, DateTime.UtcNow, null);
+
+            // Kick off the probe loop once the connection is up. Initial HostState stays
+            // Unknown until the first probe tick succeeds — avoids broadcasting a premature
+            // Running transition before any PDU round-trip has happened.
+            if (_options.Probe.Enabled)
+            {
+                _probeCts = new CancellationTokenSource();
+                _ = Task.Run(() => ProbeLoopAsync(_probeCts.Token), _probeCts.Token);
+            }
+        }
+        catch (Exception ex)
+        {
+            // Clean up a partially-constructed Plc so a retry from the caller doesn't leak
+            // the TcpClient. S7netplus's Close() is best-effort and idempotent.
+            try { Plc?.Close(); } catch { }
+            Plc = null;
+            _health = new DriverHealth(DriverState.Faulted, null, ex.Message);
+            throw;
+        }
+    }
+
+    public async Task ReinitializeAsync(string driverConfigJson, CancellationToken cancellationToken)
+    {
+        await ShutdownAsync(cancellationToken).ConfigureAwait(false);
+        await InitializeAsync(driverConfigJson, cancellationToken).ConfigureAwait(false);
+    }
+
+    public Task ShutdownAsync(CancellationToken cancellationToken)
+    {
+        try { _probeCts?.Cancel(); } catch { }
+        _probeCts?.Dispose();
+        _probeCts = null;
+
+        foreach (var state in _subscriptions.Values)
+        {
+            try { state.Cts.Cancel(); } catch { }
+            state.Cts.Dispose();
+        }
+        _subscriptions.Clear();
+
+        try { Plc?.Close(); } catch { /* best-effort — tearing down anyway */ }
+        Plc = null;
+        _health = new DriverHealth(DriverState.Unknown, _health.LastSuccessfulRead, null);
+        return Task.CompletedTask;
+    }
+
+    public DriverHealth GetHealth() => _health;
+
+    /// <summary>
+    ///     Approximate memory footprint. The Plc instance + one 240-960 byte PDU buffer is
+    ///     under 4 KB; return 0 because the <see cref="IDriver"/> contract asks for a
+    ///     driver-attributable growth number and S7.Net doesn't expose one.
+    /// </summary>
+    public long GetMemoryFootprint() => 0;
+
+    public Task FlushOptionalCachesAsync(CancellationToken cancellationToken) => Task.CompletedTask;
+
+    // ---- IReadable ----
+
+    public async Task<IReadOnlyList<DataValueSnapshot>> ReadAsync(
+        IReadOnlyList<string> fullReferences, CancellationToken cancellationToken)
+    {
+        var plc = RequirePlc();
+        var now = DateTime.UtcNow;
+        var results = new DataValueSnapshot[fullReferences.Count];
+
+        await _gate.WaitAsync(cancellationToken).ConfigureAwait(false);
+        try
+        {
+            for (var i = 0; i < fullReferences.Count; i++)
+            {
+                var name = fullReferences[i];
+                if (!_tagsByName.TryGetValue(name, out var tag))
+                {
+                    results[i] = new DataValueSnapshot(null, StatusBadNodeIdUnknown, null, now);
+                    continue;
+                }
+                try
+                {
+                    var value = await ReadOneAsync(plc, tag, cancellationToken).ConfigureAwait(false);
+                    results[i] = new DataValueSnapshot(value, 0u, now, now);
+                    _health = new DriverHealth(DriverState.Healthy, now, null);
+                }
+                catch (NotSupportedException)
+                {
+                    results[i] = new DataValueSnapshot(null, StatusBadNotSupported, null, now);
+                }
+                catch (global::S7.Net.PlcException pex)
+                {
+                    // S7.Net's PlcException carries an ErrorCode; PUT/GET-disabled on
+                    // S7-1200/1500 surfaces here. Map to BadDeviceFailure so operators see a
+                    // device-config problem (toggle PUT/GET in TIA Portal) rather than a
+                    // transient fault — per driver-specs.md §5.
+                    results[i] = new DataValueSnapshot(null, StatusBadDeviceFailure, null, now);
+                    _health = new DriverHealth(DriverState.Degraded, _health.LastSuccessfulRead, pex.Message);
+                }
+                catch (Exception ex)
+                {
+                    results[i] = new DataValueSnapshot(null, StatusBadCommunicationError, null, now);
+                    _health = new DriverHealth(DriverState.Degraded, _health.LastSuccessfulRead, ex.Message);
+                }
+            }
+        }
+        finally { _gate.Release(); }
+        return results;
+    }
+
+    private async Task<object> ReadOneAsync(global::S7.Net.Plc plc, S7TagDefinition tag, CancellationToken ct)
+    {
+        var addr = _parsedByName[tag.Name];
+        // S7.Net's string-based ReadAsync returns object where the boxed .NET type depends on
+        // the size suffix: DBX=bool, DBB=byte, DBW=ushort, DBD=uint. Our S7DataType enum
+        // specifies the SEMANTIC type (Int16 vs UInt16 vs Float32 etc.); the reinterpret below
+        // converts the raw unsigned boxed value into the requested type without issuing an
+        // extra PLC round-trip.
+        var raw = await plc.ReadAsync(tag.Address, ct).ConfigureAwait(false)
+            ?? throw new System.IO.InvalidDataException($"S7.Net returned null for '{tag.Address}'");
+
+        return (tag.DataType, addr.Size, raw) switch
+        {
+            (S7DataType.Bool, S7Size.Bit, bool b) => b,
+            (S7DataType.Byte, S7Size.Byte, byte by) => by,
+            (S7DataType.UInt16, S7Size.Word, ushort u16) => u16,
+            (S7DataType.Int16, S7Size.Word, ushort u16) => unchecked((short)u16),
+            (S7DataType.UInt32, S7Size.DWord, uint u32) => u32,
+            (S7DataType.Int32, S7Size.DWord, uint u32) => unchecked((int)u32),
+            (S7DataType.Float32, S7Size.DWord, uint u32) => BitConverter.UInt32BitsToSingle(u32),
+
+            (S7DataType.Int64, _, _)   => throw new NotSupportedException("S7 Int64 reads land in a follow-up PR"),
+            (S7DataType.UInt64, _, _)  => throw new NotSupportedException("S7 UInt64 reads land in a follow-up PR"),
+            (S7DataType.Float64, _, _) => throw new NotSupportedException("S7 Float64 (LReal) reads land in a follow-up PR"),
+            (S7DataType.String, _, _)  => throw new NotSupportedException("S7 STRING reads land in a follow-up PR"),
+            (S7DataType.DateTime, _, _) => throw new NotSupportedException("S7 DateTime reads land in a follow-up PR"),
+
+            _ => throw new System.IO.InvalidDataException(
+                $"S7 Read type-mismatch: tag '{tag.Name}' declared {tag.DataType} but address '{tag.Address}' " +
+                $"parsed as Size={addr.Size}; S7.Net returned {raw.GetType().Name}"),
+        };
+    }
+
+    // ---- IWritable ----
+
+    public async Task<IReadOnlyList<WriteResult>> WriteAsync(
+        IReadOnlyList<WriteRequest> writes, CancellationToken cancellationToken)
+    {
+        var plc = RequirePlc();
+        var results = new WriteResult[writes.Count];
+
+        await _gate.WaitAsync(cancellationToken).ConfigureAwait(false);
+        try
+        {
+            for (var i = 0; i < writes.Count; i++)
+            {
+                var w = writes[i];
+                if (!_tagsByName.TryGetValue(w.FullReference, out var tag))
+                {
+                    results[i] = new WriteResult(StatusBadNodeIdUnknown);
+                    continue;
+                }
+                if (!tag.Writable)
+                {
+                    results[i] = new WriteResult(StatusBadNotWritable);
+                    continue;
+                }
+                try
+                {
+                    await WriteOneAsync(plc, tag, w.Value, cancellationToken).ConfigureAwait(false);
+                    results[i] = new WriteResult(0u);
+                }
+                catch (NotSupportedException)
+                {
+                    results[i] = new WriteResult(StatusBadNotSupported);
+                }
+                catch (global::S7.Net.PlcException)
+                {
+                    results[i] = new WriteResult(StatusBadDeviceFailure);
+                }
+                catch (Exception)
+                {
+                    results[i] = new WriteResult(StatusBadInternalError);
+                }
+            }
+        }
+        finally { _gate.Release(); }
+        return results;
+    }
+
+    private async Task WriteOneAsync(global::S7.Net.Plc plc, S7TagDefinition tag, object? value, CancellationToken ct)
+    {
+        // S7.Net's Plc.WriteAsync(string address, object value) expects the boxed value to
+        // match the address's size-suffix type: DBX=bool, DBB=byte, DBW=ushort, DBD=uint.
+        // Our S7DataType lets the caller pass short/int/float; convert to the unsigned
+        // wire representation before handing off.
+        var boxed = tag.DataType switch
+        {
+            S7DataType.Bool    => (object)Convert.ToBoolean(value),
+            S7DataType.Byte    => (object)Convert.ToByte(value),
+            S7DataType.UInt16  => (object)Convert.ToUInt16(value),
+            S7DataType.Int16   => (object)unchecked((ushort)Convert.ToInt16(value)),
+            S7DataType.UInt32  => (object)Convert.ToUInt32(value),
+            S7DataType.Int32   => (object)unchecked((uint)Convert.ToInt32(value)),
+            S7DataType.Float32 => (object)BitConverter.SingleToUInt32Bits(Convert.ToSingle(value)),
+
+            S7DataType.Int64   => throw new NotSupportedException("S7 Int64 writes land in a follow-up PR"),
+            S7DataType.UInt64  => throw new NotSupportedException("S7 UInt64 writes land in a follow-up PR"),
+            S7DataType.Float64 => throw new NotSupportedException("S7 Float64 (LReal) writes land in a follow-up PR"),
+            S7DataType.String  => throw new NotSupportedException("S7 STRING writes land in a follow-up PR"),
+            S7DataType.DateTime => throw new NotSupportedException("S7 DateTime writes land in a follow-up PR"),
+            _ => throw new InvalidOperationException($"Unknown S7DataType {tag.DataType}"),
+        };
+        await plc.WriteAsync(tag.Address, boxed, ct).ConfigureAwait(false);
+    }
+
+    private global::S7.Net.Plc RequirePlc() =>
+        Plc ?? throw new InvalidOperationException("S7Driver not initialized");
+
+    // ---- ITagDiscovery ----
+
+    public Task DiscoverAsync(IAddressSpaceBuilder builder, CancellationToken cancellationToken)
+    {
+        ArgumentNullException.ThrowIfNull(builder);
+        var folder = builder.Folder("S7", "S7");
+        foreach (var t in _options.Tags)
+        {
+            folder.Variable(t.Name, t.Name, new DriverAttributeInfo(
+                FullName: t.Name,
+                DriverDataType: MapDataType(t.DataType),
+                IsArray: false,
+                ArrayDim: null,
+                SecurityClass: t.Writable ? SecurityClassification.Operate : SecurityClassification.ViewOnly,
+                IsHistorized: false,
+                IsAlarm: false));
+        }
+        return Task.CompletedTask;
+    }
+
+    private static DriverDataType MapDataType(S7DataType t) => t switch
+    {
+        S7DataType.Bool => DriverDataType.Boolean,
+        S7DataType.Byte => DriverDataType.Int32,      // no 8-bit in DriverDataType yet
+        S7DataType.Int16 or S7DataType.UInt16 or S7DataType.Int32 or S7DataType.UInt32 => DriverDataType.Int32,
+        S7DataType.Int64 or S7DataType.UInt64 => DriverDataType.Int32, // widens; lossy for >2^31-1
+        S7DataType.Float32 => DriverDataType.Float32,
+        S7DataType.Float64 => DriverDataType.Float64,
+        S7DataType.String => DriverDataType.String,
+        S7DataType.DateTime => DriverDataType.DateTime,
+        _ => DriverDataType.Int32,
+    };
+
+    // ---- ISubscribable (polling overlay) ----
+
+    public Task<ISubscriptionHandle> SubscribeAsync(
+        IReadOnlyList<string> fullReferences, TimeSpan publishingInterval, CancellationToken cancellationToken)
+    {
+        var id = Interlocked.Increment(ref _nextSubscriptionId);
+        var cts = new CancellationTokenSource();
+        // Floor at 100 ms — S7 CPUs scan 2-10 ms but the comms mailbox is processed at most
+        // once per scan; sub-100 ms polling just queues wire-side with worse latency.
+        var interval = publishingInterval < TimeSpan.FromMilliseconds(100)
+            ? TimeSpan.FromMilliseconds(100)
+            : publishingInterval;
+        var handle = new S7SubscriptionHandle(id);
+        var state = new SubscriptionState(handle, [.. fullReferences], interval, cts);
+        _subscriptions[id] = state;
+        _ = Task.Run(() => PollLoopAsync(state, cts.Token), cts.Token);
+        return Task.FromResult<ISubscriptionHandle>(handle);
+    }
+
+    public Task UnsubscribeAsync(ISubscriptionHandle handle, CancellationToken cancellationToken)
+    {
+        if (handle is S7SubscriptionHandle h && _subscriptions.TryRemove(h.Id, out var state))
+        {
+            state.Cts.Cancel();
+            state.Cts.Dispose();
+        }
+        return Task.CompletedTask;
+    }
+
+    private async Task PollLoopAsync(SubscriptionState state, CancellationToken ct)
+    {
+        // Initial-data push per OPC UA Part 4 convention.
+        try { await PollOnceAsync(state, forceRaise: true, ct).ConfigureAwait(false); }
+        catch (OperationCanceledException) { return; }
+        catch { /* first-read error — polling continues */ }
+
+        while (!ct.IsCancellationRequested)
+        {
+            try { await Task.Delay(state.Interval, ct).ConfigureAwait(false); }
+            catch (OperationCanceledException) { return; }
+
+            try { await PollOnceAsync(state, forceRaise: false, ct).ConfigureAwait(false); }
+            catch (OperationCanceledException) { return; }
+            catch { /* transient polling error — loop continues, health surface reflects it */ }
+        }
+    }
+
+    private async Task PollOnceAsync(SubscriptionState state, bool forceRaise, CancellationToken ct)
+    {
+        var snapshots = await ReadAsync(state.TagReferences, ct).ConfigureAwait(false);
+        for (var i = 0; i < state.TagReferences.Count; i++)
+        {
+            var tagRef = state.TagReferences[i];
+            var current = snapshots[i];
+            var lastSeen = state.LastValues.TryGetValue(tagRef, out var prev) ? prev : default;
+
+            if (forceRaise || !Equals(lastSeen?.Value, current.Value) || lastSeen?.StatusCode != current.StatusCode)
+            {
+                state.LastValues[tagRef] = current;
+                OnDataChange?.Invoke(this, new DataChangeEventArgs(state.Handle, tagRef, current));
+            }
+        }
+    }
+
+    private sealed record SubscriptionState(
+        S7SubscriptionHandle Handle,
+        IReadOnlyList<string> TagReferences,
+        TimeSpan Interval,
+        CancellationTokenSource Cts)
+    {
+        public System.Collections.Concurrent.ConcurrentDictionary<string, DataValueSnapshot> LastValues { get; }
+            = new(StringComparer.OrdinalIgnoreCase);
+    }
+
+    private sealed record S7SubscriptionHandle(long Id) : ISubscriptionHandle
+    {
+        public string DiagnosticId => $"s7-sub-{Id}";
+    }
+
+    // ---- IHostConnectivityProbe ----
+
+    /// <summary>
+    ///     Host identifier surfaced in <see cref="GetHostStatuses"/>. <c>host:port</c> format
+    ///     matches the Modbus driver's convention so the Admin UI dashboard renders both
+    ///     family's rows uniformly.
+    /// </summary>
+    public string HostName => $"{_options.Host}:{_options.Port}";
+
+    public IReadOnlyList<HostConnectivityStatus> GetHostStatuses()
+    {
+        lock (_probeLock)
+            return [new HostConnectivityStatus(HostName, _hostState, _hostStateChangedUtc)];
+    }
+
+    private async Task ProbeLoopAsync(CancellationToken ct)
+    {
+        while (!ct.IsCancellationRequested)
+        {
+            var success = false;
+            try
+            {
+                // Probe via S7.Net's low-cost GetCpuStatus — returns the CPU state (Run/Stop)
+                // and is intentionally light on the comms mailbox. Single-word Plc.ReadAsync
+                // would also work but GetCpuStatus doubles as a "PLC actually up" check.
+                using var probeCts = CancellationTokenSource.CreateLinkedTokenSource(ct);
+                probeCts.CancelAfter(_options.Probe.Timeout);
+
+                var plc = Plc;
+                if (plc is null) throw new InvalidOperationException("Plc dropped during probe");
+
+                await _gate.WaitAsync(probeCts.Token).ConfigureAwait(false);
+                try
+                {
+                    _ = await plc.ReadStatusAsync(probeCts.Token).ConfigureAwait(false);
+                    success = true;
+                }
+                finally { _gate.Release(); }
+            }
+            catch (OperationCanceledException) when (ct.IsCancellationRequested) { return; }
+            catch { /* transport/timeout/exception — treated as Stopped below */ }
+
+            TransitionTo(success ? HostState.Running : HostState.Stopped);
+
+            try { await Task.Delay(_options.Probe.Interval, ct).ConfigureAwait(false); }
+            catch (OperationCanceledException) { return; }
+        }
+    }
+
+    private void TransitionTo(HostState newState)
+    {
+        HostState old;
+        lock (_probeLock)
+        {
+            old = _hostState;
+            if (old == newState) return;
+            _hostState = newState;
+            _hostStateChangedUtc = DateTime.UtcNow;
+        }
+        OnHostStatusChanged?.Invoke(this, new HostStatusChangedEventArgs(HostName, old, newState));
+    }
+
+    public void Dispose() => DisposeAsync().AsTask().GetAwaiter().GetResult();
+
+    public async ValueTask DisposeAsync()
+    {
+        if (_disposed) return;
+        _disposed = true;
+        try { await ShutdownAsync(CancellationToken.None).ConfigureAwait(false); }
+        catch { /* disposal is best-effort */ }
+        _gate.Dispose();
+    }
+}

src/ZB.MOM.WW.OtOpcUa.Driver.S7/S7DriverOptions.cs

@@ -0,0 +1,112 @@
+using S7NetCpuType = global::S7.Net.CpuType;
+
+namespace ZB.MOM.WW.OtOpcUa.Driver.S7;
+
+/// <summary>
+///     Siemens S7 native (S7comm / ISO-on-TCP port 102) driver configuration. Bound from the
+///     driver's <c>DriverConfig</c> JSON at <c>DriverHost.RegisterAsync</c>. Unlike the Modbus
+///     driver the S7 driver uses the PLC's *native* protocol — port 102 ISO-on-TCP rather
+///     than Modbus's 502, and S7-specific area codes (DB, M, I, Q) rather than holding-
+///     register / coil tables.
+/// </summary>
+/// <remarks>
+///     <para>
+///         The driver requires <b>PUT/GET communication enabled</b> in the TIA Portal
+///         hardware config for S7-1200/1500. The factory default disables PUT/GET access,
+///         so a driver configured against a freshly-flashed CPU will see a hard error
+///         (S7.Net surfaces it as <c>Plc.ReadAsync</c> returning <c>ErrorCode.Accessing</c>).
+///         The driver maps that specifically to <c>BadNotSupported</c> and flags it as a
+///         configuration alert rather than a transient fault — blind Polly retry is wasted
+///         effort when the PLC will keep refusing every request.
+///     </para>
+///     <para>
+///         See <c>docs/v2/driver-specs.md</c> §5 for the full specification.
+///     </para>
+/// </remarks>
+public sealed class S7DriverOptions
+{
+    /// <summary>PLC IP address or hostname.</summary>
+    public string Host { get; init; } = "127.0.0.1";
+
+    /// <summary>TCP port. ISO-on-TCP is 102 on every S7 model; override only for unusual NAT setups.</summary>
+    public int Port { get; init; } = 102;
+
+    /// <summary>
+    ///     CPU family. Determines the ISO-TSAP slot byte that S7.Net uses during connection
+    ///     setup — pick the family that matches the target PLC exactly.
+    /// </summary>
+    public S7NetCpuType CpuType { get; init; } = S7NetCpuType.S71500;
+
+    /// <summary>
+    ///     Hardware rack number. Almost always 0; relevant only for distributed S7-400 racks
+    ///     with multiple CPUs.
+    /// </summary>
+    public short Rack { get; init; } = 0;
+
+    /// <summary>
+    ///     CPU slot. Conventions per family: S7-300 = slot 2, S7-400 = slot 2 or 3,
+    ///     S7-1200 / S7-1500 = slot 0 (onboard PN). S7.Net uses this to build the remote
+    ///     TSAP. Wrong slot → connection refused during handshake.
+    /// </summary>
+    public short Slot { get; init; } = 0;
+
+    /// <summary>Connect + per-operation timeout.</summary>
+    public TimeSpan Timeout { get; init; } = TimeSpan.FromSeconds(5);
+
+    /// <summary>Pre-declared tag map. S7 has a symbol-table protocol but S7.Net does not expose it, so the driver operates off a static tag list configured per-site. Address grammar documented in S7AddressParser (PR 63).</summary>
+    public IReadOnlyList<S7TagDefinition> Tags { get; init; } = [];
+
+    /// <summary>
+    ///     Background connectivity-probe settings. When enabled, the driver runs a tick loop
+    ///     that issues a cheap read against <see cref="S7ProbeOptions.ProbeAddress"/> every
+    ///     <see cref="S7ProbeOptions.Interval"/> and raises <c>OnHostStatusChanged</c> on
+    ///     Running ↔ Stopped transitions.
+    /// </summary>
+    public S7ProbeOptions Probe { get; init; } = new();
+}
+
+public sealed class S7ProbeOptions
+{
+    public bool Enabled { get; init; } = true;
+    public TimeSpan Interval { get; init; } = TimeSpan.FromSeconds(5);
+    public TimeSpan Timeout { get; init; } = TimeSpan.FromSeconds(2);
+
+    /// <summary>
+    ///     Address to probe for liveness. DB1.DBW0 is the convention if the PLC project
+    ///     reserves a small fingerprint DB for health checks (per <c>docs/v2/s7.md</c>);
+    ///     if not, pick any valid Merker word like <c>MW0</c>.
+    /// </summary>
+    public string ProbeAddress { get; init; } = "MW0";
+}
+
+/// <summary>
+///     One S7 variable as exposed by the driver. Addresses use S7.Net syntax — see
+///     <c>S7AddressParser</c> (PR 63) for the grammar.
+/// </summary>
+/// <param name="Name">Tag name; OPC UA browse name + driver full reference.</param>
+/// <param name="Address">S7 address string, e.g. <c>DB1.DBW0</c>, <c>M0.0</c>, <c>I0.0</c>, <c>QD4</c>. Grammar documented in <c>S7AddressParser</c> (PR 63).</param>
+/// <param name="DataType">Logical data type — drives the underlying S7.Net read/write width.</param>
+/// <param name="Writable">When true the driver accepts writes for this tag.</param>
+/// <param name="StringLength">For <c>DataType = String</c>: S7-string max length. Default 254 (S7 max).</param>
+public sealed record S7TagDefinition(
+    string Name,
+    string Address,
+    S7DataType DataType,
+    bool Writable = true,
+    int StringLength = 254);
+
+public enum S7DataType
+{
+    Bool,
+    Byte,
+    Int16,
+    UInt16,
+    Int32,
+    UInt32,
+    Int64,
+    UInt64,
+    Float32,
+    Float64,
+    String,
+    DateTime,
+}

src/ZB.MOM.WW.OtOpcUa.Driver.S7/ZB.MOM.WW.OtOpcUa.Driver.S7.csproj

@@ -0,0 +1,27 @@
+<Project Sdk="Microsoft.NET.Sdk">
+
+    <PropertyGroup>
+        <TargetFramework>net10.0</TargetFramework>
+        <Nullable>enable</Nullable>
+        <ImplicitUsings>enable</ImplicitUsings>
+        <LangVersion>latest</LangVersion>
+        <TreatWarningsAsErrors>true</TreatWarningsAsErrors>
+        <GenerateDocumentationFile>true</GenerateDocumentationFile>
+        <NoWarn>$(NoWarn);CS1591</NoWarn>
+        <RootNamespace>ZB.MOM.WW.OtOpcUa.Driver.S7</RootNamespace>
+        <AssemblyName>ZB.MOM.WW.OtOpcUa.Driver.S7</AssemblyName>
+    </PropertyGroup>
+
+    <ItemGroup>
+        <ProjectReference Include="..\ZB.MOM.WW.OtOpcUa.Core.Abstractions\ZB.MOM.WW.OtOpcUa.Core.Abstractions.csproj"/>
+    </ItemGroup>
+
+    <ItemGroup>
+        <PackageReference Include="S7netplus" Version="0.20.0"/>
+    </ItemGroup>
+
+    <ItemGroup>
+        <InternalsVisibleTo Include="ZB.MOM.WW.OtOpcUa.Driver.S7.Tests"/>
+    </ItemGroup>
+
+</Project>

tests/ZB.MOM.WW.OtOpcUa.Driver.Modbus.IntegrationTests/Mitsubishi/MitsubishiProfile.cs

@@ -0,0 +1,43 @@
+namespace ZB.MOM.WW.OtOpcUa.Driver.Modbus.IntegrationTests.Mitsubishi;
+
+/// <summary>
+///     Tag map for the Mitsubishi MELSEC device class with a representative Modbus Device
+///     Assignment block mapping D0..D1023 → HR[0..1023]. Mirrors the behaviors in
+///     <c>mitsubishi.json</c> pymodbus profile and <c>docs/v2/mitsubishi.md</c>.
+/// </summary>
+/// <remarks>
+///     MELSEC Modbus sites all have *different* device-assignment parameter blocks; this profile
+///     models the conventional default. Per-model differences (FX5U needs firmware ≥ 1.060 for
+///     Modbus server; QJ71MT91 lacks FC22/FC23; FX/iQ-F use octal X/Y while Q/L/iQ-R use hex)
+///     are handled in <see cref="MelsecAddress"/> (PR 59) and the per-model test files.
+/// </remarks>
+public static class MitsubishiProfile
+{
+    /// <summary>
+    ///     Scratch HR the smoke test writes + reads. Address 200 mirrors the
+    ///     dl205/s7_1500/standard scratch range so one smoke test pattern works across every
+    ///     device profile the simulator supports.
+    /// </summary>
+    public const ushort SmokeHoldingRegister = 200;
+
+    /// <summary>Value the smoke test writes then reads back.</summary>
+    public const short SmokeHoldingValue = 7890;
+
+    public static ModbusDriverOptions BuildOptions(string host, int port) => new()
+    {
+        Host = host,
+        Port = port,
+        UnitId = 1,
+        Timeout = TimeSpan.FromSeconds(2),
+        Tags =
+        [
+            new ModbusTagDefinition(
+                Name: "Smoke_HReg200",
+                Region: ModbusRegion.HoldingRegisters,
+                Address: SmokeHoldingRegister,
+                DataType: ModbusDataType.Int16,
+                Writable: true),
+        ],
+        Probe = new ModbusProbeOptions { Enabled = false },
+    };
+}

tests/ZB.MOM.WW.OtOpcUa.Driver.Modbus.IntegrationTests/Mitsubishi/MitsubishiQuirkTests.cs

@@ -0,0 +1,179 @@
+using Shouldly;
+using Xunit;
+
+namespace ZB.MOM.WW.OtOpcUa.Driver.Modbus.IntegrationTests.Mitsubishi;
+
+/// <summary>
+///     Verifies the MELSEC-family Modbus quirks against the <c>mitsubishi.json</c> pymodbus
+///     profile: CDAB word order default, binary-not-BCD D-register encoding, hex X-input
+///     parsing (Q/L/iQ-R), D0 fingerprint, M-relay coil mapping with bank base.
+/// </summary>
+/// <remarks>
+///     Groups all quirks in one test class instead of per-behavior classes (unlike the DL205
+///     set) because MELSEC's per-model differentiation is handled by the
+///     <see cref="MelsecFamily"/> enum on the helper + <c>MODBUS_SIM_PROFILE</c> env var on
+///     the fixture, rather than per-PR test classes.
+/// </remarks>
+[Collection(ModbusSimulatorCollection.Name)]
+[Trait("Category", "Integration")]
+[Trait("Device", "Mitsubishi")]
+public sealed class MitsubishiQuirkTests(ModbusSimulatorFixture sim)
+{
+    [Fact]
+    public async Task Mitsubishi_D0_fingerprint_reads_0x1234()
+    {
+        if (!ShouldRun()) return;
+        await using var driver = await NewDriverAsync(
+            new ModbusTagDefinition("D0_Fingerprint",
+                ModbusRegion.HoldingRegisters,
+                Address: MelsecAddress.DRegisterToHolding("D0"),
+                DataType: ModbusDataType.UInt16, Writable: false));
+
+        var r = await driver.ReadAsync(["D0_Fingerprint"], TestContext.Current.CancellationToken);
+        r[0].StatusCode.ShouldBe(0u);
+        r[0].Value.ShouldBe((ushort)0x1234);
+    }
+
+    [Fact]
+    public async Task Mitsubishi_Float32_CDAB_decodes_1_5f_from_D100()
+    {
+        if (!ShouldRun()) return;
+        // MELSEC Q/L/iQ-R/iQ-F all store 32-bit values with CDAB word order (low word at
+        // lower D-register address). HR[100..101] = [0, 0x3FC0] decodes as 1.5f under
+        // WordSwap but as a denormal under BigEndian.
+        var addr = MelsecAddress.DRegisterToHolding("D100");
+        await using var driver = await NewDriverAsync(
+            new ModbusTagDefinition("D100_Float_CDAB",
+                ModbusRegion.HoldingRegisters, Address: addr,
+                DataType: ModbusDataType.Float32, Writable: false,
+                ByteOrder: ModbusByteOrder.WordSwap),
+            new ModbusTagDefinition("D100_Float_ABCD_control",
+                ModbusRegion.HoldingRegisters, Address: addr,
+                DataType: ModbusDataType.Float32, Writable: false,
+                ByteOrder: ModbusByteOrder.BigEndian));
+
+        var r = await driver.ReadAsync(
+            ["D100_Float_CDAB", "D100_Float_ABCD_control"],
+            TestContext.Current.CancellationToken);
+        r[0].Value.ShouldBe(1.5f, "MELSEC stores Float32 CDAB; WordSwap decode returns 1.5f");
+        r[1].Value.ShouldNotBe(1.5f, "same wire with BigEndian must decode to a different value");
+    }
+
+    [Fact]
+    public async Task Mitsubishi_D10_is_binary_not_BCD()
+    {
+        if (!ShouldRun()) return;
+        // Counter-to-DL205: MELSEC D-registers are binary by default. D10 = 1234 decimal =
+        // 0x04D2. Reading as Int16 returns 1234; reading as Bcd16 would throw (nibble 0xD is
+        // non-BCD) — the integration test proves the Int16 decode wins.
+        await using var driver = await NewDriverAsync(
+            new ModbusTagDefinition("D10_Binary",
+                ModbusRegion.HoldingRegisters,
+                Address: MelsecAddress.DRegisterToHolding("D10"),
+                DataType: ModbusDataType.Int16, Writable: false));
+
+        var r = await driver.ReadAsync(["D10_Binary"], TestContext.Current.CancellationToken);
+        r[0].StatusCode.ShouldBe(0u);
+        r[0].Value.ShouldBe((short)1234, "MELSEC stores numeric D-register values in binary; 0x04D2 = 1234");
+    }
+
+    [Fact]
+    public async Task Mitsubishi_D10_as_BCD_throws_because_nibble_is_non_decimal()
+    {
+        if (!ShouldRun()) return;
+        // If a site configured D10 with Bcd16 data type but the ladder writes binary, the
+        // BCD decoder MUST reject the garbage rather than silently returning wrong decimal.
+        // 0x04D2 contains nibble 0xD which fails BCD validation.
+        await using var driver = await NewDriverAsync(
+            new ModbusTagDefinition("D10_WrongBcd",
+                ModbusRegion.HoldingRegisters,
+                Address: MelsecAddress.DRegisterToHolding("D10"),
+                DataType: ModbusDataType.Bcd16, Writable: false));
+
+        var r = await driver.ReadAsync(["D10_WrongBcd"], TestContext.Current.CancellationToken);
+        // ReadAsync catches the InvalidDataException from DecodeBcd and surfaces it as
+        // BadCommunicationError (PR 52 mapping). Non-zero status = caller sees a real
+        // problem and can check their tag config instead of getting silently-wrong numbers.
+        r[0].StatusCode.ShouldNotBe(0u, "BCD decode of binary 0x04D2 must fail loudly because nibble D is non-BCD");
+    }
+
+    [Fact]
+    public async Task Mitsubishi_QLiQR_X210_hex_maps_to_DI_528_reads_ON()
+    {
+        if (!ShouldRun()) return;
+        // MELSEC-Q / L / iQ-R: X addresses are hex. X210 = 0x210 = 528 decimal.
+        // mitsubishi.json seeds cell 33 (DI 528..543) with value 9 = bit 0 + bit 3 set.
+        // X210 → DI 528 → cell 33 bit 0 = 1 (ON).
+        var addr = MelsecAddress.XInputToDiscrete("X210", MelsecFamily.Q_L_iQR);
+        addr.ShouldBe((ushort)528);
+
+        await using var driver = await NewDriverAsync(
+            new ModbusTagDefinition("X210_hex",
+                ModbusRegion.DiscreteInputs, Address: addr,
+                DataType: ModbusDataType.Bool, Writable: false));
+
+        var r = await driver.ReadAsync(["X210_hex"], TestContext.Current.CancellationToken);
+        r[0].StatusCode.ShouldBe(0u);
+        r[0].Value.ShouldBe(true);
+    }
+
+    [Fact]
+    public void Mitsubishi_family_trap_X20_differs_on_Q_vs_FX()
+    {
+        // Not a live-sim test — a unit-level proof that the MELSEC family selector gates the
+        // address correctly. Included in the integration suite so anyone running the MELSEC
+        // tests sees the trap called out explicitly.
+        MelsecAddress.XInputToDiscrete("X20", MelsecFamily.Q_L_iQR).ShouldBe((ushort)32);
+        MelsecAddress.XInputToDiscrete("X20", MelsecFamily.F_iQF).ShouldBe((ushort)16);
+    }
+
+    [Fact]
+    public async Task Mitsubishi_M512_maps_to_coil_512_reads_ON()
+    {
+        if (!ShouldRun()) return;
+        // mitsubishi.json seeds cell 32 (coil 512..527) with value 5 = bit 0 + bit 2 set.
+        // M512 → coil 512 → cell 32 bit 0 = 1 (ON).
+        var addr = MelsecAddress.MRelayToCoil("M512");
+        addr.ShouldBe((ushort)512);
+
+        await using var driver = await NewDriverAsync(
+            new ModbusTagDefinition("M512",
+                ModbusRegion.Coils, Address: addr,
+                DataType: ModbusDataType.Bool, Writable: false));
+
+        var r = await driver.ReadAsync(["M512"], TestContext.Current.CancellationToken);
+        r[0].StatusCode.ShouldBe(0u);
+        r[0].Value.ShouldBe(true);
+    }
+
+    // --- helpers ---
+
+    private bool ShouldRun()
+    {
+        if (sim.SkipReason is not null) { Assert.Skip(sim.SkipReason); return false; }
+        if (!string.Equals(Environment.GetEnvironmentVariable("MODBUS_SIM_PROFILE"), "mitsubishi",
+                StringComparison.OrdinalIgnoreCase))
+        {
+            Assert.Skip("MODBUS_SIM_PROFILE != mitsubishi — skipping.");
+            return false;
+        }
+        return true;
+    }
+
+    private async Task<ModbusDriver> NewDriverAsync(params ModbusTagDefinition[] tags)
+    {
+        var drv = new ModbusDriver(
+            new ModbusDriverOptions
+            {
+                Host = sim.Host,
+                Port = sim.Port,
+                UnitId = 1,
+                Timeout = TimeSpan.FromSeconds(2),
+                Tags = tags,
+                Probe = new ModbusProbeOptions { Enabled = false },
+            },
+            driverInstanceId: "melsec-quirk");
+        await drv.InitializeAsync("{}", TestContext.Current.CancellationToken);
+        return drv;
+    }
+}

tests/ZB.MOM.WW.OtOpcUa.Driver.Modbus.IntegrationTests/Mitsubishi/MitsubishiSmokeTests.cs

@@ -0,0 +1,45 @@
+using Shouldly;
+using Xunit;
+
+namespace ZB.MOM.WW.OtOpcUa.Driver.Modbus.IntegrationTests.Mitsubishi;
+
+/// <summary>
+///     End-to-end smoke against the MELSEC <c>mitsubishi.json</c> pymodbus profile (or a real
+///     MELSEC QJ71MT91 / iQ-R / FX5U when <c>MODBUS_SIM_ENDPOINT</c> points at one). Drives
+///     the full <see cref="ModbusDriver"/> + real <see cref="ModbusTcpTransport"/> stack.
+///     Success proves the driver initializes against the MELSEC sim, writes a known value,
+///     and reads it back — the baseline every Mitsubishi-specific test (PR 59+) builds on.
+/// </summary>
+[Collection(ModbusSimulatorCollection.Name)]
+[Trait("Category", "Integration")]
+[Trait("Device", "Mitsubishi")]
+public sealed class MitsubishiSmokeTests(ModbusSimulatorFixture sim)
+{
+    [Fact]
+    public async Task Mitsubishi_roundtrip_write_then_read_of_holding_register()
+    {
+        if (sim.SkipReason is not null) Assert.Skip(sim.SkipReason);
+        if (!string.Equals(Environment.GetEnvironmentVariable("MODBUS_SIM_PROFILE"), "mitsubishi",
+                StringComparison.OrdinalIgnoreCase))
+        {
+            Assert.Skip("MODBUS_SIM_PROFILE != mitsubishi — skipping.");
+        }
+
+        var options = MitsubishiProfile.BuildOptions(sim.Host, sim.Port);
+        await using var driver = new ModbusDriver(options, driverInstanceId: "melsec-smoke");
+        await driver.InitializeAsync(driverConfigJson: "{}", TestContext.Current.CancellationToken);
+
+        var writeResults = await driver.WriteAsync(
+            [new(FullReference: "Smoke_HReg200", Value: (short)MitsubishiProfile.SmokeHoldingValue)],
+            TestContext.Current.CancellationToken);
+        writeResults.Count.ShouldBe(1);
+        writeResults[0].StatusCode.ShouldBe(0u, "write must succeed against the MELSEC pymodbus profile");
+
+        var readResults = await driver.ReadAsync(
+            ["Smoke_HReg200"],
+            TestContext.Current.CancellationToken);
+        readResults.Count.ShouldBe(1);
+        readResults[0].StatusCode.ShouldBe(0u);
+        readResults[0].Value.ShouldBe((short)MitsubishiProfile.SmokeHoldingValue);
+    }
+}

tests/ZB.MOM.WW.OtOpcUa.Driver.Modbus.IntegrationTests/Pymodbus/mitsubishi.json

@@ -0,0 +1,83 @@
+{
+  "_comment": "mitsubishi.json -- Mitsubishi MELSEC Modbus TCP quirk simulator covering QJ71MT91, iQ-R, iQ-F/FX5U, and FX3U-ENET-P502 behaviors documented in docs/v2/mitsubishi.md. MELSEC CPUs store multi-word values in CDAB order (opposite of S7 ABCD, same family as DL260). The Modbus-module 'Modbus Device Assignment Parameter' block is per-site, so this profile models one *representative* assignment mapping D-register D0..D1023 -> HR 0..1023, M-relay M0..M511 -> coil 0..511, X-input X0..X15 -> DI 0..15 (X-addresses are HEX on Q/L/iQ-R, so X10 = decimal 16; on FX/iQ-F they're OCTAL like DL260). pymodbus bit-address semantics are the same as dl205.json and s7_1500.json (FC01/02/05/15 address N maps to cell index N/16).",
+
+  "server_list": {
+    "srv": {
+      "comm": "tcp",
+      "host": "0.0.0.0",
+      "port": 5020,
+      "framer": "socket",
+      "device_id": 1
+    }
+  },
+
+  "device_list": {
+    "dev": {
+      "setup": {
+        "co size":    4096,
+        "di size":    4096,
+        "hr size":    4096,
+        "ir size":    1024,
+        "shared blocks": true,
+        "type exception": false,
+        "defaults": {
+          "value":  {"bits": 0, "uint16": 0, "uint32": 0, "float32": 0.0, "string": " "},
+          "action": {"bits": null, "uint16": null, "uint32": null, "float32": null, "string": null}
+        }
+      },
+      "invalid": [],
+      "write": [
+        [0, 0],
+        [10, 10],
+        [100, 101],
+        [200, 209],
+        [300, 301],
+        [500, 500]
+      ],
+
+      "uint16": [
+        {"_quirk": "D0 fingerprint marker. MELSEC D0 is the first data register; Modbus Device Assignment typically maps D0..D1023 -> HR 0..1023. 0x1234 is the fingerprint operators set in GX Works to prove the mapping parameter block is in effect.",
+         "addr": 0, "value": 4660},
+
+        {"_quirk": "Scratch HR range 200..209 -- mirrors the dl205/s7_1500/standard scratch range so smoke tests (MitsubishiProfile.SmokeHoldingRegister=200) round-trip identically against any profile.",
+         "addr": 200, "value": 0},
+        {"addr": 201, "value": 0},
+        {"addr": 202, "value": 0},
+        {"addr": 203, "value": 0},
+        {"addr": 204, "value": 0},
+        {"addr": 205, "value": 0},
+        {"addr": 206, "value": 0},
+        {"addr": 207, "value": 0},
+        {"addr": 208, "value": 0},
+        {"addr": 209, "value": 0},
+
+        {"_quirk": "Float32 1.5f in CDAB word order (MELSEC Q/L/iQ-R/iQ-F default, same as DL260). HR[100]=0x0000=0 low word, HR[101]=0x3FC0=16320 high word. Decode with ByteOrder.WordSwap returns 1.5f; BigEndian decode returns a denormal.",
+         "addr": 100, "value": 0},
+        {"addr": 101, "value": 16320},
+
+        {"_quirk": "Int32 0x12345678 in CDAB word order. HR[300]=0x5678=22136 low word, HR[301]=0x1234=4660 high word. Contrasts with the S7 profile's ABCD encoding at the same address.",
+         "addr": 300, "value": 22136},
+        {"addr": 301, "value": 4660},
+
+        {"_quirk": "D10 = decimal 1234 stored as BINARY (NOT BCD like DL205). 0x04D2 = 1234 decimal. Caller reading with Bcd16 data type would decode this as binary 1234's BCD nibbles which are non-BCD and throw InvalidDataException -- proves MELSEC is binary-by-default, opposite of DL205's BCD-by-default quirk.",
+         "addr": 10, "value": 1234},
+
+        {"_quirk": "Modbus Device Assignment boundary marker. HR[500] represents the last register in an assigned D-range D500. Beyond this (HR[501..4095]) would be Illegal Data Address on a real QJ71MT91 with this specific parameter block; pymodbus returns default 0 because its shared cell array has space -- real-PLC parity is documented in docs/v2/mitsubishi.md §device-assignment, not enforced here.",
+         "addr": 500, "value": 500}
+      ],
+
+      "bits": [
+        {"_quirk": "M-relay marker cell at cell 32 = Modbus coil 512 = MELSEC M512 (coils 0..15 collide with the D0 uint16 marker cell, so we place the M marker above that). Cell 32 bit 0 = 1 and bit 2 = 1 (value = 0b101 = 5) = M512=ON, M513=OFF, M514=ON. Matches the Y0/Y2 marker pattern in dl205 and s7_1500 profiles.",
+         "addr": 32, "value": 5},
+
+        {"_quirk": "X-input marker cell at cell 33 = Modbus DI 528 (= MELSEC X210 hex on Q/L/iQ-R). Cell 33 bit 0 = 1 and bit 3 = 1 (value = 0x9 = 9). Chosen above cell 1 so it doesn't collide with any uint16 D-register. Proves the hex-parsing X-input helper on Q/L/iQ-R family; FX/iQ-F families use octal X-addresses tested separately.",
+         "addr": 33, "value": 9}
+      ],
+
+      "uint32":  [],
+      "float32": [],
+      "string":  [],
+      "repeat":  []
+    }
+  }
+}

tests/ZB.MOM.WW.OtOpcUa.Driver.Modbus.IntegrationTests/Pymodbus/s7_1500.json

@@ -0,0 +1,77 @@
+{
+  "_comment": "s7_1500.json -- Siemens SIMATIC S7-1500 + MB_SERVER quirk simulator. Models docs/v2/s7.md behaviors as concrete register values. Unlike DL260 (CDAB word order default) or Mitsubishi (CDAB default), S7 MB_SERVER uses ABCD word order by default because Siemens native CPU types are big-endian top-to-bottom both within the register pair and byte pair. This profile exists so the driver's S7 profile default ByteOrder.BigEndian can be validated end-to-end. pymodbus bit-address semantics are the same as dl205.json (FC01/02/05/15 address X maps to cell index X/16); seed bits at the appropriate cell-indexed positions.",
+
+  "server_list": {
+    "srv": {
+      "comm": "tcp",
+      "host": "0.0.0.0",
+      "port": 5020,
+      "framer": "socket",
+      "device_id": 1
+    }
+  },
+
+  "device_list": {
+    "dev": {
+      "setup": {
+        "co size":    4096,
+        "di size":    4096,
+        "hr size":    4096,
+        "ir size":    1024,
+        "shared blocks": true,
+        "type exception": false,
+        "defaults": {
+          "value":  {"bits": 0, "uint16": 0, "uint32": 0, "float32": 0.0, "string": " "},
+          "action": {"bits": null, "uint16": null, "uint32": null, "float32": null, "string": null}
+        }
+      },
+      "invalid": [],
+      "write": [
+        [0, 0],
+        [25, 25],
+        [100, 101],
+        [200, 209],
+        [300, 301]
+      ],
+
+      "uint16": [
+        {"_quirk": "DB1 header marker. On an S7-1500 with MB_SERVER pointing at DB1, operators often reserve DB1.DBW0 for a fingerprint word so clients can verify they're talking to the right DB. 0xABCD = 43981.",
+         "addr": 0, "value": 43981},
+
+        {"_quirk": "Scratch HR range 200..209 -- mirrors the standard.json scratch range so the smoke test (S7_1500Profile.SmokeHoldingRegister=200) round-trips identically against either profile.",
+         "addr": 200, "value": 0},
+        {"addr": 201, "value": 0},
+        {"addr": 202, "value": 0},
+        {"addr": 203, "value": 0},
+        {"addr": 204, "value": 0},
+        {"addr": 205, "value": 0},
+        {"addr": 206, "value": 0},
+        {"addr": 207, "value": 0},
+        {"addr": 208, "value": 0},
+        {"addr": 209, "value": 0},
+
+        {"_quirk": "Float32 1.5f in ABCD word order (Siemens big-endian default, OPPOSITE of DL260 CDAB). IEEE-754 1.5 = 0x3FC00000. ABCD = high word first: HR[100]=0x3FC0=16320, HR[101]=0x0000=0.",
+         "addr": 100, "value": 16320},
+        {"_quirk": "Float32 1.5f ABCD low word.",
+         "addr": 101, "value": 0},
+
+        {"_quirk": "Int32 0x12345678 in ABCD word order. HR[300]=0x1234=4660, HR[301]=0x5678=22136. Demonstrates the contrast with DL260 CDAB Int32 encoding.",
+         "addr": 300, "value": 4660},
+        {"addr": 301, "value": 22136}
+      ],
+
+      "bits": [
+        {"_quirk": "Coil bank marker cell. S7 MB_SERVER doesn't fix coil addresses; this simulates a user-wired DB where coil 400 (=bit 0 of cell 25) represents a latched digital output. Cell 25 bit 0 = 1 proves the wire-format round-trip works for coils on S7 profile.",
+         "addr": 25, "value": 1},
+
+        {"_quirk": "Discrete-input bank marker cell. DI 500 (=bit 0 of cell 31) = 1. Like coils, discrete inputs on S7 MB_SERVER are per-site; we assert the end-to-end FC02 path only.",
+         "addr": 31, "value": 1}
+      ],
+
+      "uint32":  [],
+      "float32": [],
+      "string":  [],
+      "repeat":  []
+    }
+  }
+}

tests/ZB.MOM.WW.OtOpcUa.Driver.Modbus.IntegrationTests/Pymodbus/serve.ps1

@@ -21,7 +21,7 @@
 #>
 [CmdletBinding()]
 param(
-    [Parameter(Mandatory)] [ValidateSet('standard', 'dl205')] [string]$Profile,
+    [Parameter(Mandatory)] [ValidateSet('standard', 'dl205', 's7_1500', 'mitsubishi')] [string]$Profile,
     [int]$HttpPort = 8080
 )
 

tests/ZB.MOM.WW.OtOpcUa.Driver.Modbus.IntegrationTests/S7/S7_1500Profile.cs

@@ -0,0 +1,44 @@
+namespace ZB.MOM.WW.OtOpcUa.Driver.Modbus.IntegrationTests.S7;
+
+/// <summary>
+///     Tag map for the Siemens SIMATIC S7-1500 device class with the <c>MB_SERVER</c> library
+///     block mapping HR[0..] to DB1.DBW0+. Mirrors <c>s7_1500.json</c> in <c>Pymodbus/</c>.
+/// </summary>
+/// <remarks>
+///     Unlike DL205, S7 has no fixed Modbus memory map — every site wires MB_SERVER to a
+///     different DB. The profile here models the *default* user layout documented in
+///     <c>docs/v2/s7.md</c> §per-model-matrix: DB1.DBW0 = fingerprint marker, a scratch HR
+///     range 200..209 for write-roundtrip tests, and ABCD-order Float32 / Int32 markers at
+///     HR[100..101] and HR[300..301] to prove the driver's S7 profile default is correct.
+/// </remarks>
+public static class S7_1500Profile
+{
+    /// <summary>
+    ///     Scratch HR the smoke test writes + reads. Address 200 mirrors the DL205 /
+    ///     standard scratch range so one smoke test pattern works across all device profiles.
+    /// </summary>
+    public const ushort SmokeHoldingRegister = 200;
+
+    /// <summary>Value the smoke test writes then reads back.</summary>
+    public const short SmokeHoldingValue = 4321;
+
+    public static ModbusDriverOptions BuildOptions(string host, int port) => new()
+    {
+        Host = host,
+        Port = port,
+        UnitId = 1,
+        Timeout = TimeSpan.FromSeconds(2),
+        Tags =
+        [
+            new ModbusTagDefinition(
+                Name: "Smoke_HReg200",
+                Region: ModbusRegion.HoldingRegisters,
+                Address: SmokeHoldingRegister,
+                DataType: ModbusDataType.Int16,
+                Writable: true),
+        ],
+        // Disable the background probe loop — integration tests drive reads explicitly and
+        // the probe would race with assertions.
+        Probe = new ModbusProbeOptions { Enabled = false },
+    };
+}

tests/ZB.MOM.WW.OtOpcUa.Driver.Modbus.IntegrationTests/S7/S7_1500SmokeTests.cs

@@ -0,0 +1,54 @@
+using Shouldly;
+using Xunit;
+
+namespace ZB.MOM.WW.OtOpcUa.Driver.Modbus.IntegrationTests.S7;
+
+/// <summary>
+///     End-to-end smoke against the S7-1500 <c>MB_SERVER</c> pymodbus profile (or a real
+///     S7-1500 + MB_SERVER deployment when <c>MODBUS_SIM_ENDPOINT</c> points at one). Drives
+///     the full <see cref="ModbusDriver"/> + real <see cref="ModbusTcpTransport"/> stack —
+///     no fake transport. Success proves the driver initializes against the S7 simulator,
+///     writes a known value, and reads it back with the correct status and value, which is
+///     the baseline every S7-specific test (PR 57+) builds on.
+/// </summary>
+/// <remarks>
+///     S7-specific quirk tests (MB_SERVER requires non-optimized DBs, ABCD word order
+///     default, port-per-connection, FC23 Illegal Function, STOP-mode behaviour, etc.) land
+///     as separate test classes in this directory as each quirk is validated in pymodbus.
+///     Keep this smoke test deliberately narrow — filtering by device class
+///     (<c>--filter DisplayName~S7</c>) should surface the quirk-specific failure mode when
+///     something goes wrong, not a blanket smoke failure that could mean anything.
+/// </remarks>
+[Collection(ModbusSimulatorCollection.Name)]
+[Trait("Category", "Integration")]
+[Trait("Device", "S7")]
+public sealed class S7_1500SmokeTests(ModbusSimulatorFixture sim)
+{
+    [Fact]
+    public async Task S7_1500_roundtrip_write_then_read_of_holding_register()
+    {
+        if (sim.SkipReason is not null) Assert.Skip(sim.SkipReason);
+        if (!string.Equals(Environment.GetEnvironmentVariable("MODBUS_SIM_PROFILE"), "s7_1500",
+                StringComparison.OrdinalIgnoreCase))
+        {
+            Assert.Skip("MODBUS_SIM_PROFILE != s7_1500 — skipping (other profiles don't seed the S7 scratch range identically).");
+        }
+
+        var options = S7_1500Profile.BuildOptions(sim.Host, sim.Port);
+        await using var driver = new ModbusDriver(options, driverInstanceId: "s7-smoke");
+        await driver.InitializeAsync(driverConfigJson: "{}", TestContext.Current.CancellationToken);
+
+        var writeResults = await driver.WriteAsync(
+            [new(FullReference: "Smoke_HReg200", Value: (short)S7_1500Profile.SmokeHoldingValue)],
+            TestContext.Current.CancellationToken);
+        writeResults.Count.ShouldBe(1);
+        writeResults[0].StatusCode.ShouldBe(0u, "write must succeed against the S7-1500 MB_SERVER profile");
+
+        var readResults = await driver.ReadAsync(
+            ["Smoke_HReg200"],
+            TestContext.Current.CancellationToken);
+        readResults.Count.ShouldBe(1);
+        readResults[0].StatusCode.ShouldBe(0u);
+        readResults[0].Value.ShouldBe((short)S7_1500Profile.SmokeHoldingValue);
+    }
+}

tests/ZB.MOM.WW.OtOpcUa.Driver.Modbus.IntegrationTests/S7/S7_ByteOrderTests.cs

@@ -0,0 +1,132 @@
+using Shouldly;
+using Xunit;
+
+namespace ZB.MOM.WW.OtOpcUa.Driver.Modbus.IntegrationTests.S7;
+
+/// <summary>
+///     Verifies the Siemens S7 big-endian (<c>ABCD</c>) word-order default for Float32 and
+///     Int32 against the <c>s7_1500.json</c> pymodbus profile. S7's native CPU types are
+///     big-endian end-to-end, so <c>MB_SERVER</c> places the high word at the lower register
+///     address — <b>opposite</b> of DL260's CDAB. The driver's S7-family tag config must
+///     therefore default to <see cref="ModbusByteOrder.BigEndian"/>; selecting
+///     <see cref="ModbusByteOrder.WordSwap"/> against an S7 would decode garbage.
+/// </summary>
+[Collection(ModbusSimulatorCollection.Name)]
+[Trait("Category", "Integration")]
+[Trait("Device", "S7")]
+public sealed class S7_ByteOrderTests(ModbusSimulatorFixture sim)
+{
+    [Fact]
+    public async Task S7_Float32_ABCD_decodes_1_5f_from_HR100()
+    {
+        if (sim.SkipReason is not null) Assert.Skip(sim.SkipReason);
+        if (!string.Equals(Environment.GetEnvironmentVariable("MODBUS_SIM_PROFILE"), "s7_1500",
+                StringComparison.OrdinalIgnoreCase))
+        {
+            Assert.Skip("MODBUS_SIM_PROFILE != s7_1500 — skipping (s7_1500 profile is the only one seeding HR[100..101] ABCD).");
+        }
+
+        var options = new ModbusDriverOptions
+        {
+            Host = sim.Host,
+            Port = sim.Port,
+            UnitId = 1,
+            Timeout = TimeSpan.FromSeconds(2),
+            Tags =
+            [
+                new ModbusTagDefinition("S7_Float_ABCD",
+                    ModbusRegion.HoldingRegisters, Address: 100,
+                    DataType: ModbusDataType.Float32, Writable: false,
+                    ByteOrder: ModbusByteOrder.BigEndian),
+                // Control: same address with WordSwap should decode garbage — proves the
+                // two code paths diverge on S7 wire bytes.
+                new ModbusTagDefinition("S7_Float_CDAB_control",
+                    ModbusRegion.HoldingRegisters, Address: 100,
+                    DataType: ModbusDataType.Float32, Writable: false,
+                    ByteOrder: ModbusByteOrder.WordSwap),
+            ],
+            Probe = new ModbusProbeOptions { Enabled = false },
+        };
+        await using var driver = new ModbusDriver(options, driverInstanceId: "s7-float-abcd");
+        await driver.InitializeAsync("{}", TestContext.Current.CancellationToken);
+
+        var results = await driver.ReadAsync(
+            ["S7_Float_ABCD", "S7_Float_CDAB_control"],
+            TestContext.Current.CancellationToken);
+
+        results[0].StatusCode.ShouldBe(0u);
+        results[0].Value.ShouldBe(1.5f, "S7 MB_SERVER stores Float32 in ABCD word order; BigEndian decode returns 1.5f");
+
+        results[1].StatusCode.ShouldBe(0u);
+        results[1].Value.ShouldNotBe(1.5f, "applying CDAB swap to S7 ABCD bytes must produce a different value — confirms the flag is not a no-op and S7 profile default must be BigEndian");
+    }
+
+    [Fact]
+    public async Task S7_Int32_ABCD_decodes_0x12345678_from_HR300()
+    {
+        if (sim.SkipReason is not null) Assert.Skip(sim.SkipReason);
+        if (!string.Equals(Environment.GetEnvironmentVariable("MODBUS_SIM_PROFILE"), "s7_1500",
+                StringComparison.OrdinalIgnoreCase))
+        {
+            Assert.Skip("MODBUS_SIM_PROFILE != s7_1500 — skipping.");
+        }
+
+        var options = new ModbusDriverOptions
+        {
+            Host = sim.Host,
+            Port = sim.Port,
+            UnitId = 1,
+            Timeout = TimeSpan.FromSeconds(2),
+            Tags =
+            [
+                new ModbusTagDefinition("S7_Int32_ABCD",
+                    ModbusRegion.HoldingRegisters, Address: 300,
+                    DataType: ModbusDataType.Int32, Writable: false,
+                    ByteOrder: ModbusByteOrder.BigEndian),
+            ],
+            Probe = new ModbusProbeOptions { Enabled = false },
+        };
+        await using var driver = new ModbusDriver(options, driverInstanceId: "s7-int-abcd");
+        await driver.InitializeAsync("{}", TestContext.Current.CancellationToken);
+
+        var results = await driver.ReadAsync(["S7_Int32_ABCD"], TestContext.Current.CancellationToken);
+        results[0].StatusCode.ShouldBe(0u);
+        results[0].Value.ShouldBe(0x12345678,
+            "S7 Int32 stored as HR[300]=0x1234, HR[301]=0x5678 with ABCD order decodes to 0x12345678 — DL260 would store the reverse order");
+    }
+
+    [Fact]
+    public async Task S7_DB1_fingerprint_marker_at_HR0_reads_0xABCD()
+    {
+        if (sim.SkipReason is not null) Assert.Skip(sim.SkipReason);
+        if (!string.Equals(Environment.GetEnvironmentVariable("MODBUS_SIM_PROFILE"), "s7_1500",
+                StringComparison.OrdinalIgnoreCase))
+        {
+            Assert.Skip("MODBUS_SIM_PROFILE != s7_1500 — skipping.");
+        }
+
+        // Real-world MB_SERVER deployments typically reserve DB1.DBW0 as a fingerprint so
+        // clients can verify they're pointing at the right DB (protects against typos in
+        // the MB_SERVER.MB_HOLD_REG.DB_number parameter). 0xABCD is the convention.
+        var options = new ModbusDriverOptions
+        {
+            Host = sim.Host,
+            Port = sim.Port,
+            UnitId = 1,
+            Timeout = TimeSpan.FromSeconds(2),
+            Tags =
+            [
+                new ModbusTagDefinition("S7_Fingerprint",
+                    ModbusRegion.HoldingRegisters, Address: 0,
+                    DataType: ModbusDataType.UInt16, Writable: false),
+            ],
+            Probe = new ModbusProbeOptions { Enabled = false },
+        };
+        await using var driver = new ModbusDriver(options, driverInstanceId: "s7-fingerprint");
+        await driver.InitializeAsync("{}", TestContext.Current.CancellationToken);
+
+        var results = await driver.ReadAsync(["S7_Fingerprint"], TestContext.Current.CancellationToken);
+        results[0].StatusCode.ShouldBe(0u);
+        results[0].Value.ShouldBe((ushort)0xABCD);
+    }
+}

tests/ZB.MOM.WW.OtOpcUa.Driver.Modbus.IntegrationTests/ZB.MOM.WW.OtOpcUa.Driver.Modbus.IntegrationTests.csproj

@@ -26,6 +26,8 @@
     <ItemGroup>
         <None Update="Pymodbus\**\*" CopyToOutputDirectory="PreserveNewest"/>
         <None Update="DL205\**\*" CopyToOutputDirectory="PreserveNewest"/>
+        <None Update="S7\**\*" CopyToOutputDirectory="PreserveNewest"/>
+        <None Update="Mitsubishi\**\*" CopyToOutputDirectory="PreserveNewest"/>
     </ItemGroup>
 
     <ItemGroup>

tests/ZB.MOM.WW.OtOpcUa.Driver.Modbus.Tests/MelsecAddressTests.cs

@@ -0,0 +1,116 @@
+using Shouldly;
+using Xunit;
+
+namespace ZB.MOM.WW.OtOpcUa.Driver.Modbus.Tests;
+
+[Trait("Category", "Unit")]
+public sealed class MelsecAddressTests
+{
+    // --- X / Y hex vs octal family trap ---
+
+    [Theory]
+    [InlineData("X0", (ushort)0)]
+    [InlineData("X9", (ushort)9)]
+    [InlineData("XA", (ushort)10)]      // hex
+    [InlineData("XF", (ushort)15)]
+    [InlineData("X10", (ushort)16)]     // hex 0x10 = decimal 16
+    [InlineData("X20", (ushort)32)]     // hex 0x20 = decimal 32 — the classic MELSEC-Q trap
+    [InlineData("X1FF", (ushort)511)]
+    [InlineData("x10", (ushort)16)]     // lowercase prefix
+    public void XInputToDiscrete_QLiQR_parses_hex(string x, ushort expected)
+        => MelsecAddress.XInputToDiscrete(x, MelsecFamily.Q_L_iQR).ShouldBe(expected);
+
+    [Theory]
+    [InlineData("X0", (ushort)0)]
+    [InlineData("X7", (ushort)7)]
+    [InlineData("X10", (ushort)8)]      // octal 10 = decimal 8
+    [InlineData("X20", (ushort)16)]     // octal 20 = decimal 16 — SAME string, DIFFERENT value on FX
+    [InlineData("X777", (ushort)511)]
+    public void XInputToDiscrete_FiQF_parses_octal(string x, ushort expected)
+        => MelsecAddress.XInputToDiscrete(x, MelsecFamily.F_iQF).ShouldBe(expected);
+
+    [Theory]
+    [InlineData("Y0", (ushort)0)]
+    [InlineData("Y1F", (ushort)31)]
+    public void YOutputToCoil_QLiQR_parses_hex(string y, ushort expected)
+        => MelsecAddress.YOutputToCoil(y, MelsecFamily.Q_L_iQR).ShouldBe(expected);
+
+    [Theory]
+    [InlineData("Y0", (ushort)0)]
+    [InlineData("Y17", (ushort)15)]
+    public void YOutputToCoil_FiQF_parses_octal(string y, ushort expected)
+        => MelsecAddress.YOutputToCoil(y, MelsecFamily.F_iQF).ShouldBe(expected);
+
+    [Fact]
+    public void Same_address_string_decodes_differently_between_families()
+    {
+        // This is the headline quirk: "X20" in GX Works means one thing on Q-series and
+        // another on FX-series. The driver's family selector is the only defence.
+        MelsecAddress.XInputToDiscrete("X20", MelsecFamily.Q_L_iQR).ShouldBe((ushort)32);
+        MelsecAddress.XInputToDiscrete("X20", MelsecFamily.F_iQF).ShouldBe((ushort)16);
+    }
+
+    [Theory]
+    [InlineData("X8")]     // 8 is non-octal
+    [InlineData("X12G")]   // G is non-hex
+    public void XInputToDiscrete_FiQF_rejects_non_octal(string bad)
+        => Should.Throw<ArgumentException>(() => MelsecAddress.XInputToDiscrete(bad, MelsecFamily.F_iQF));
+
+    [Theory]
+    [InlineData("X12G")]
+    public void XInputToDiscrete_QLiQR_rejects_non_hex(string bad)
+        => Should.Throw<ArgumentException>(() => MelsecAddress.XInputToDiscrete(bad, MelsecFamily.Q_L_iQR));
+
+    [Fact]
+    public void XInputToDiscrete_honors_bank_base_from_assignment_block()
+    {
+        // Real-world QJ71MT91 assignment blocks commonly place X at DI 8192+ when other
+        // ranges take the low Modbus addresses. Helper must add the base cleanly.
+        MelsecAddress.XInputToDiscrete("X10", MelsecFamily.Q_L_iQR, xBankBase: 8192).ShouldBe((ushort)(8192 + 16));
+    }
+
+    // --- M-relay (decimal, both families) ---
+
+    [Theory]
+    [InlineData("M0", (ushort)0)]
+    [InlineData("M10", (ushort)10)]     // M addresses are DECIMAL, not hex or octal
+    [InlineData("M511", (ushort)511)]
+    [InlineData("m99", (ushort)99)]     // lowercase
+    public void MRelayToCoil_parses_decimal(string m, ushort expected)
+        => MelsecAddress.MRelayToCoil(m).ShouldBe(expected);
+
+    [Fact]
+    public void MRelayToCoil_honors_bank_base()
+        => MelsecAddress.MRelayToCoil("M0", mBankBase: 512).ShouldBe((ushort)512);
+
+    [Fact]
+    public void MRelayToCoil_rejects_non_numeric()
+        => Should.Throw<ArgumentException>(() => MelsecAddress.MRelayToCoil("M1F"));
+
+    // --- D-register (decimal, both families) ---
+
+    [Theory]
+    [InlineData("D0", (ushort)0)]
+    [InlineData("D100", (ushort)100)]
+    [InlineData("d1023", (ushort)1023)]
+    public void DRegisterToHolding_parses_decimal(string d, ushort expected)
+        => MelsecAddress.DRegisterToHolding(d).ShouldBe(expected);
+
+    [Fact]
+    public void DRegisterToHolding_honors_bank_base()
+        => MelsecAddress.DRegisterToHolding("D10", dBankBase: 4096).ShouldBe((ushort)4106);
+
+    [Fact]
+    public void DRegisterToHolding_rejects_empty()
+        => Should.Throw<ArgumentException>(() => MelsecAddress.DRegisterToHolding("D"));
+
+    // --- overflow ---
+
+    [Fact]
+    public void XInputToDiscrete_overflow_throws()
+    {
+        // 0xFFFF + base 1 = 0x10000 — past ushort.
+        Should.Throw<OverflowException>(() =>
+            MelsecAddress.XInputToDiscrete("XFFFF", MelsecFamily.Q_L_iQR, xBankBase: 1));
+    }
+}

tests/ZB.MOM.WW.OtOpcUa.Driver.OpcUaClient.Tests/OpcUaClientDriverScaffoldTests.cs

@@ -0,0 +1,90 @@
+using Shouldly;
+using Xunit;
+using ZB.MOM.WW.OtOpcUa.Core.Abstractions;
+
+namespace ZB.MOM.WW.OtOpcUa.Driver.OpcUaClient.Tests;
+
+/// <summary>
+///     Scaffold-level tests for <see cref="OpcUaClientDriver"/> that don't require a live
+///     remote OPC UA server. PR 67+ adds IReadable/IWritable/ITagDiscovery/ISubscribable
+///     tests against a local in-process OPC UA server fixture.
+/// </summary>
+[Trait("Category", "Unit")]
+public sealed class OpcUaClientDriverScaffoldTests
+{
+    [Fact]
+    public void Default_options_target_standard_opcua_port_and_anonymous_auth()
+    {
+        var opts = new OpcUaClientDriverOptions();
+        opts.EndpointUrl.ShouldBe("opc.tcp://localhost:4840", "4840 is the IANA-assigned OPC UA port");
+        opts.SecurityMode.ShouldBe(OpcUaSecurityMode.None);
+        opts.AuthType.ShouldBe(OpcUaAuthType.Anonymous);
+        opts.AutoAcceptCertificates.ShouldBeFalse("production default must reject untrusted server certs");
+    }
+
+    [Fact]
+    public void Default_timeouts_match_driver_specs_section_8()
+    {
+        var opts = new OpcUaClientDriverOptions();
+        opts.SessionTimeout.ShouldBe(TimeSpan.FromSeconds(120));
+        opts.KeepAliveInterval.ShouldBe(TimeSpan.FromSeconds(5));
+        opts.ReconnectPeriod.ShouldBe(TimeSpan.FromSeconds(5));
+    }
+
+    [Fact]
+    public void Driver_reports_type_and_id_before_connect()
+    {
+        using var drv = new OpcUaClientDriver(new OpcUaClientDriverOptions(), "opcua-test");
+        drv.DriverType.ShouldBe("OpcUaClient");
+        drv.DriverInstanceId.ShouldBe("opcua-test");
+        drv.GetHealth().State.ShouldBe(DriverState.Unknown);
+    }
+
+    [Fact]
+    public async Task Initialize_against_unreachable_endpoint_transitions_to_Faulted_and_throws()
+    {
+        // RFC 5737 reserved-for-documentation IP; won't route anywhere. Pick opc.tcp:// so
+        // endpoint selection hits the transport-layer connection rather than a DNS lookup.
+        var opts = new OpcUaClientDriverOptions
+        {
+            // Port 1 on loopback is effectively guaranteed to be closed — the OS responds
+            // with TCP RST immediately instead of hanging on connect, which keeps the
+            // unreachable-host tests snappy. Don't use an RFC 5737 reserved IP; those get
+            // routed to a black-hole + time out only after the SDK's internal retry/backoff
+            // fully elapses (~60s even with Options.Timeout=500ms).
+            EndpointUrl = "opc.tcp://127.0.0.1:1",
+            Timeout = TimeSpan.FromMilliseconds(500),
+            AutoAcceptCertificates = true, // dev-mode to bypass cert validation in the test
+        };
+        using var drv = new OpcUaClientDriver(opts, "opcua-unreach");
+
+        await Should.ThrowAsync<Exception>(async () =>
+            await drv.InitializeAsync("{}", TestContext.Current.CancellationToken));
+
+        var health = drv.GetHealth();
+        health.State.ShouldBe(DriverState.Faulted);
+        health.LastError.ShouldNotBeNull();
+    }
+
+    [Fact]
+    public async Task Reinitialize_against_unreachable_endpoint_re_throws()
+    {
+        var opts = new OpcUaClientDriverOptions
+        {
+            // Port 1 on loopback is effectively guaranteed to be closed — the OS responds
+            // with TCP RST immediately instead of hanging on connect, which keeps the
+            // unreachable-host tests snappy. Don't use an RFC 5737 reserved IP; those get
+            // routed to a black-hole + time out only after the SDK's internal retry/backoff
+            // fully elapses (~60s even with Options.Timeout=500ms).
+            EndpointUrl = "opc.tcp://127.0.0.1:1",
+            Timeout = TimeSpan.FromMilliseconds(500),
+            AutoAcceptCertificates = true,
+        };
+        using var drv = new OpcUaClientDriver(opts, "opcua-reinit");
+
+        await Should.ThrowAsync<Exception>(async () =>
+            await drv.InitializeAsync("{}", TestContext.Current.CancellationToken));
+        await Should.ThrowAsync<Exception>(async () =>
+            await drv.ReinitializeAsync("{}", TestContext.Current.CancellationToken));
+    }
+}

tests/ZB.MOM.WW.OtOpcUa.Driver.OpcUaClient.Tests/OpcUaClientReadWriteTests.cs

@@ -0,0 +1,32 @@
+using Shouldly;
+using Xunit;
+using ZB.MOM.WW.OtOpcUa.Core.Abstractions;
+
+namespace ZB.MOM.WW.OtOpcUa.Driver.OpcUaClient.Tests;
+
+/// <summary>
+///     Unit tests for the IReadable/IWritable surface that don't need a live remote OPC UA
+///     server. Wire-level round-trips against a local in-process server fixture land in a
+///     follow-up PR once we have one scaffolded.
+/// </summary>
+[Trait("Category", "Unit")]
+public sealed class OpcUaClientReadWriteTests
+{
+    [Fact]
+    public async Task ReadAsync_without_initialize_throws_InvalidOperationException()
+    {
+        using var drv = new OpcUaClientDriver(new OpcUaClientDriverOptions(), "opcua-uninit");
+        await Should.ThrowAsync<InvalidOperationException>(async () =>
+            await drv.ReadAsync(["ns=2;s=Demo"], TestContext.Current.CancellationToken));
+    }
+
+    [Fact]
+    public async Task WriteAsync_without_initialize_throws_InvalidOperationException()
+    {
+        using var drv = new OpcUaClientDriver(new OpcUaClientDriverOptions(), "opcua-uninit");
+        await Should.ThrowAsync<InvalidOperationException>(async () =>
+            await drv.WriteAsync(
+                [new WriteRequest("ns=2;s=Demo", 42)],
+                TestContext.Current.CancellationToken));
+    }
+}

tests/ZB.MOM.WW.OtOpcUa.Driver.OpcUaClient.Tests/ZB.MOM.WW.OtOpcUa.Driver.OpcUaClient.Tests.csproj

@@ -0,0 +1,31 @@
+<Project Sdk="Microsoft.NET.Sdk">
+
+    <PropertyGroup>
+        <TargetFramework>net10.0</TargetFramework>
+        <Nullable>enable</Nullable>
+        <ImplicitUsings>enable</ImplicitUsings>
+        <IsPackable>false</IsPackable>
+        <IsTestProject>true</IsTestProject>
+        <RootNamespace>ZB.MOM.WW.OtOpcUa.Driver.OpcUaClient.Tests</RootNamespace>
+    </PropertyGroup>
+
+    <ItemGroup>
+        <PackageReference Include="xunit.v3" Version="1.1.0"/>
+        <PackageReference Include="Shouldly" Version="4.3.0"/>
+        <PackageReference Include="Microsoft.NET.Test.Sdk" Version="17.12.0"/>
+        <PackageReference Include="xunit.runner.visualstudio" Version="3.0.2">
+            <PrivateAssets>all</PrivateAssets>
+            <IncludeAssets>runtime; build; native; contentfiles; analyzers; buildtransitive</IncludeAssets>
+        </PackageReference>
+    </ItemGroup>
+
+    <ItemGroup>
+        <ProjectReference Include="..\..\src\ZB.MOM.WW.OtOpcUa.Driver.OpcUaClient\ZB.MOM.WW.OtOpcUa.Driver.OpcUaClient.csproj"/>
+    </ItemGroup>
+
+    <ItemGroup>
+        <NuGetAuditSuppress Include="https://github.com/advisories/GHSA-37gx-xxp4-5rgx"/>
+        <NuGetAuditSuppress Include="https://github.com/advisories/GHSA-w3x6-4m5h-cxqf"/>
+    </ItemGroup>
+
+</Project>

tests/ZB.MOM.WW.OtOpcUa.Driver.S7.Tests/S7AddressParserTests.cs

@@ -0,0 +1,119 @@
+using Shouldly;
+using Xunit;
+
+namespace ZB.MOM.WW.OtOpcUa.Driver.S7.Tests;
+
+[Trait("Category", "Unit")]
+public sealed class S7AddressParserTests
+{
+    // --- Data blocks ---
+
+    [Theory]
+    [InlineData("DB1.DBX0.0", 1, S7Size.Bit, 0, 0)]
+    [InlineData("DB1.DBX0.7", 1, S7Size.Bit, 0, 7)]
+    [InlineData("DB1.DBB0", 1, S7Size.Byte, 0, 0)]
+    [InlineData("DB1.DBW0", 1, S7Size.Word, 0, 0)]
+    [InlineData("DB1.DBD4", 1, S7Size.DWord, 4, 0)]
+    [InlineData("DB10.DBW100", 10, S7Size.Word, 100, 0)]
+    [InlineData("DB1.DBX15.3", 1, S7Size.Bit, 15, 3)]
+    public void Parse_data_block_addresses(string input, int db, S7Size size, int byteOff, int bitOff)
+    {
+        var r = S7AddressParser.Parse(input);
+        r.Area.ShouldBe(S7Area.DataBlock);
+        r.DbNumber.ShouldBe(db);
+        r.Size.ShouldBe(size);
+        r.ByteOffset.ShouldBe(byteOff);
+        r.BitOffset.ShouldBe(bitOff);
+    }
+
+    [Theory]
+    [InlineData("db1.dbw0", 1, S7Size.Word, 0)]
+    [InlineData(" DB1.DBW0 ", 1, S7Size.Word, 0)]       // trim whitespace
+    public void Parse_is_case_insensitive_and_trims(string input, int db, S7Size size, int off)
+    {
+        var r = S7AddressParser.Parse(input);
+        r.Area.ShouldBe(S7Area.DataBlock);
+        r.DbNumber.ShouldBe(db);
+        r.Size.ShouldBe(size);
+        r.ByteOffset.ShouldBe(off);
+    }
+
+    // --- M / I / Q ---
+
+    [Theory]
+    [InlineData("MB0", S7Area.Memory, S7Size.Byte, 0, 0)]
+    [InlineData("MW10", S7Area.Memory, S7Size.Word, 10, 0)]
+    [InlineData("MD4", S7Area.Memory, S7Size.DWord, 4, 0)]
+    [InlineData("M0.0", S7Area.Memory, S7Size.Bit, 0, 0)]
+    [InlineData("M255.7", S7Area.Memory, S7Size.Bit, 255, 7)]
+    [InlineData("IB0", S7Area.Input, S7Size.Byte, 0, 0)]
+    [InlineData("IW0", S7Area.Input, S7Size.Word, 0, 0)]
+    [InlineData("I0.0", S7Area.Input, S7Size.Bit, 0, 0)]
+    [InlineData("QB0", S7Area.Output, S7Size.Byte, 0, 0)]
+    [InlineData("QW0", S7Area.Output, S7Size.Word, 0, 0)]
+    [InlineData("Q0.0", S7Area.Output, S7Size.Bit, 0, 0)]
+    [InlineData("QD4", S7Area.Output, S7Size.DWord, 4, 0)]
+    public void Parse_MIQ_addresses(string input, S7Area area, S7Size size, int byteOff, int bitOff)
+    {
+        var r = S7AddressParser.Parse(input);
+        r.Area.ShouldBe(area);
+        r.DbNumber.ShouldBe(0);
+        r.Size.ShouldBe(size);
+        r.ByteOffset.ShouldBe(byteOff);
+        r.BitOffset.ShouldBe(bitOff);
+    }
+
+    // --- Timers / counters ---
+
+    [Theory]
+    [InlineData("T0", S7Area.Timer, 0)]
+    [InlineData("T15", S7Area.Timer, 15)]
+    [InlineData("C0", S7Area.Counter, 0)]
+    [InlineData("C10", S7Area.Counter, 10)]
+    public void Parse_timer_and_counter(string input, S7Area area, int number)
+    {
+        var r = S7AddressParser.Parse(input);
+        r.Area.ShouldBe(area);
+        r.ByteOffset.ShouldBe(number);
+        r.Size.ShouldBe(S7Size.Word, "timers + counters are 16-bit opaque");
+    }
+
+    // --- Reject garbage ---
+
+    [Theory]
+    [InlineData("")]
+    [InlineData("   ")]
+    [InlineData("Z0")]                 // unknown area
+    [InlineData("DB")]                 // no number or tail
+    [InlineData("DB1")]                // no tail
+    [InlineData("DB1.")]               // empty tail
+    [InlineData("DB1.DBX0")]           // bit size without .bit
+    [InlineData("DB1.DBX0.8")]         // bit 8 out of range
+    [InlineData("DB1.DBW0.0")]         // word with bit suffix
+    [InlineData("DB0.DBW0")]           // db 0 invalid
+    [InlineData("DBA.DBW0")]           // non-numeric db
+    [InlineData("DB1.DBQ0")]           // invalid size letter
+    [InlineData("M")]                  // no offset
+    [InlineData("M0")]                 // bit access needs .bit
+    [InlineData("M0.8")]               // bit 8
+    [InlineData("MB-1")]               // negative offset
+    [InlineData("MW")]                 // no offset digits
+    [InlineData("TA")]                 // non-numeric timer
+    public void Parse_rejects_invalid(string bad)
+        => Should.Throw<FormatException>(() => S7AddressParser.Parse(bad));
+
+    [Fact]
+    public void TryParse_returns_false_for_garbage_without_throwing()
+    {
+        S7AddressParser.TryParse("not-an-address", out var r).ShouldBeFalse();
+        r.ShouldBe(default);
+    }
+
+    [Fact]
+    public void TryParse_returns_true_for_valid_address()
+    {
+        S7AddressParser.TryParse("DB1.DBW0", out var r).ShouldBeTrue();
+        r.DbNumber.ShouldBe(1);
+        r.Size.ShouldBe(S7Size.Word);
+    }
+}

tests/ZB.MOM.WW.OtOpcUa.Driver.S7.Tests/S7DiscoveryAndSubscribeTests.cs

@@ -0,0 +1,117 @@
+using Shouldly;
+using Xunit;
+using ZB.MOM.WW.OtOpcUa.Core.Abstractions;
+
+namespace ZB.MOM.WW.OtOpcUa.Driver.S7.Tests;
+
+/// <summary>
+///     Shape tests for <see cref="S7Driver"/>'s <see cref="ITagDiscovery"/>,
+///     <see cref="ISubscribable"/>, and <see cref="IHostConnectivityProbe"/> surfaces that
+///     don't need a live PLC. Wire-level polling round-trips and probe transitions land in a
+///     follow-up PR once we have a mock S7 server.
+/// </summary>
+[Trait("Category", "Unit")]
+public sealed class S7DiscoveryAndSubscribeTests
+{
+    private sealed class RecordingAddressSpaceBuilder : IAddressSpaceBuilder
+    {
+        public readonly List<string> Folders = new();
+        public readonly List<(string Name, DriverAttributeInfo Attr)> Variables = new();
+
+        public IAddressSpaceBuilder Folder(string browseName, string displayName)
+        {
+            Folders.Add(browseName);
+            return this;
+        }
+        public IVariableHandle Variable(string browseName, string displayName, DriverAttributeInfo attributeInfo)
+        {
+            Variables.Add((browseName, attributeInfo));
+            return new StubHandle();
+        }
+        public void AddProperty(string browseName, DriverDataType dataType, object? value) { }
+        public void AttachAlarmCondition(IVariableHandle sourceVariable, string alarmName, DriverAttributeInfo alarmInfo) { }
+
+        private sealed class StubHandle : IVariableHandle
+        {
+            public string FullReference => "stub";
+            public IAlarmConditionSink MarkAsAlarmCondition(AlarmConditionInfo info)
+                => throw new NotImplementedException("S7 driver never calls this — no alarm surfacing");
+        }
+    }
+
+    [Fact]
+    public async Task DiscoverAsync_projects_every_tag_into_the_address_space()
+    {
+        var opts = new S7DriverOptions
+        {
+            Host = "192.0.2.1",
+            Tags =
+            [
+                new("TempSetpoint", "DB1.DBW0", S7DataType.Int16, Writable: true),
+                new("FaultBit", "M0.0", S7DataType.Bool, Writable: false),
+                new("PIDOutput", "DB5.DBD12", S7DataType.Float32, Writable: true),
+            ],
+        };
+        using var drv = new S7Driver(opts, "s7-disco");
+
+        var builder = new RecordingAddressSpaceBuilder();
+        await drv.DiscoverAsync(builder, TestContext.Current.CancellationToken);
+
+        builder.Folders.ShouldContain("S7");
+        builder.Variables.Count.ShouldBe(3);
+        builder.Variables[0].Name.ShouldBe("TempSetpoint");
+        builder.Variables[0].Attr.SecurityClass.ShouldBe(SecurityClassification.Operate, "writable tags get Operate security class");
+        builder.Variables[1].Attr.SecurityClass.ShouldBe(SecurityClassification.ViewOnly, "read-only tags get ViewOnly");
+        builder.Variables[2].Attr.DriverDataType.ShouldBe(DriverDataType.Float32);
+    }
+
+    [Fact]
+    public void GetHostStatuses_returns_one_row_with_host_port_identity_pre_init()
+    {
+        var opts = new S7DriverOptions { Host = "plc1.internal", Port = 102 };
+        using var drv = new S7Driver(opts, "s7-host");
+
+        var rows = drv.GetHostStatuses();
+        rows.Count.ShouldBe(1);
+        rows[0].HostName.ShouldBe("plc1.internal:102");
+        rows[0].State.ShouldBe(HostState.Unknown, "pre-init / pre-probe state is Unknown");
+    }
+
+    [Fact]
+    public async Task SubscribeAsync_returns_unique_handles_and_UnsubscribeAsync_accepts_them()
+    {
+        var opts = new S7DriverOptions { Host = "192.0.2.1" };
+        using var drv = new S7Driver(opts, "s7-sub");
+
+        // SubscribeAsync does not itself call ReadAsync (the poll task does), so this works
+        // even though the driver isn't initialized. The poll task catches the resulting
+        // InvalidOperationException and the loop quietly continues — same pattern as the
+        // Modbus driver's poll loop tolerating transient transport failures.
+        var h1 = await drv.SubscribeAsync(["T1"], TimeSpan.FromMilliseconds(200), TestContext.Current.CancellationToken);
+        var h2 = await drv.SubscribeAsync(["T2"], TimeSpan.FromMilliseconds(200), TestContext.Current.CancellationToken);
+
+        h1.DiagnosticId.ShouldStartWith("s7-sub-");
+        h2.DiagnosticId.ShouldStartWith("s7-sub-");
+        h1.DiagnosticId.ShouldNotBe(h2.DiagnosticId);
+
+        await drv.UnsubscribeAsync(h1, TestContext.Current.CancellationToken);
+        await drv.UnsubscribeAsync(h2, TestContext.Current.CancellationToken);
+        // UnsubscribeAsync with an unknown handle must be a no-op, not throw.
+        await drv.UnsubscribeAsync(h1, TestContext.Current.CancellationToken);
+    }
+
+    [Fact]
+    public async Task Subscribe_publishing_interval_is_floored_at_100ms()
+    {
+        var opts = new S7DriverOptions { Host = "192.0.2.1", Probe = new S7ProbeOptions { Enabled = false } };
+        using var drv = new S7Driver(opts, "s7-floor");
+
+        // 50 ms requested — the floor protects the S7 CPU from sub-scan polling that would
+        // just queue wire-side. Test that the subscription is accepted (the floor is applied
+        // internally; the floor value isn't exposed, so we're really just asserting that the
+        // driver doesn't reject small intervals).
+        var h = await drv.SubscribeAsync(["T"], TimeSpan.FromMilliseconds(50), TestContext.Current.CancellationToken);
+        h.ShouldNotBeNull();
+        await drv.UnsubscribeAsync(h, TestContext.Current.CancellationToken);
+    }
+}

tests/ZB.MOM.WW.OtOpcUa.Driver.S7.Tests/S7DriverReadWriteTests.cs

@@ -0,0 +1,54 @@
+using Shouldly;
+using Xunit;
+
+namespace ZB.MOM.WW.OtOpcUa.Driver.S7.Tests;
+
+/// <summary>
+///     Unit tests for <see cref="S7Driver"/>'s <c>IReadable</c>/<c>IWritable</c> surface
+///     that don't require a live PLC — covers error paths (not-initialized, unknown tag,
+///     read-only write rejection, unsupported data types). Wire-level round-trip tests
+///     against a live S7 or a mock-server land in a follow-up PR since S7.Net doesn't ship
+///     an in-process fake and an adequate mock is non-trivial.
+/// </summary>
+[Trait("Category", "Unit")]
+public sealed class S7DriverReadWriteTests
+{
+    [Fact]
+    public async Task Initialize_rejects_invalid_tag_address_and_fails_fast()
+    {
+        // Bad address at init time must throw; the alternative (deferring the parse to the
+        // first read) would surface the config bug as BadInternalError on every subsequent
+        // Read which is impossible for an operator to diagnose from the OPC UA client.
+        var opts = new S7DriverOptions
+        {
+            Host = "192.0.2.1", // reserved — will never complete TCP handshake
+            Timeout = TimeSpan.FromMilliseconds(250),
+            Tags = [new S7TagDefinition("BadTag", "NOT-AN-S7-ADDRESS", S7DataType.Int16)],
+        };
+        using var drv = new S7Driver(opts, "s7-bad-tag");
+
+        // Either the TCP connect fails first (Exception) or the parser fails (FormatException)
+        // — both are acceptable since both are init-time fail-fast. What matters is that we
+        // don't return a "healthy" driver with a latent bad tag.
+        await Should.ThrowAsync<Exception>(async () =>
+            await drv.InitializeAsync("{}", TestContext.Current.CancellationToken));
+    }
+
+    [Fact]
+    public async Task ReadAsync_without_initialize_throws_InvalidOperationException()
+    {
+        using var drv = new S7Driver(new S7DriverOptions { Host = "192.0.2.1" }, "s7-uninit");
+        await Should.ThrowAsync<InvalidOperationException>(async () =>
+            await drv.ReadAsync(["Any"], TestContext.Current.CancellationToken));
+    }
+
+    [Fact]
+    public async Task WriteAsync_without_initialize_throws_InvalidOperationException()
+    {
+        using var drv = new S7Driver(new S7DriverOptions { Host = "192.0.2.1" }, "s7-uninit");
+        await Should.ThrowAsync<InvalidOperationException>(async () =>
+            await drv.WriteAsync(
+                [new(FullReference: "Any", Value: (short)0)],
+                TestContext.Current.CancellationToken));
+    }
+}

tests/ZB.MOM.WW.OtOpcUa.Driver.S7.Tests/S7DriverScaffoldTests.cs

@@ -0,0 +1,66 @@
+using Shouldly;
+using Xunit;
+using ZB.MOM.WW.OtOpcUa.Core.Abstractions;
+
+namespace ZB.MOM.WW.OtOpcUa.Driver.S7.Tests;
+
+/// <summary>
+///     Scaffold-level tests that don't need a live S7 PLC — exercise driver lifecycle shape,
+///     default option values, and failure-mode transitions. PR 64 adds IReadable/IWritable
+///     tests against a mock-server, PR 65 adds discovery + subscribe.
+/// </summary>
+[Trait("Category", "Unit")]
+public sealed class S7DriverScaffoldTests
+{
+    [Fact]
+    public void Default_options_target_S7_1500_slot_0_on_port_102()
+    {
+        var opts = new S7DriverOptions();
+        opts.Port.ShouldBe(102, "ISO-on-TCP is always 102 for S7; documented in driver-specs.md §5");
+        opts.CpuType.ShouldBe(global::S7.Net.CpuType.S71500);
+        opts.Rack.ShouldBe((short)0);
+        opts.Slot.ShouldBe((short)0, "S7-1200/1500 onboard PN ports are slot 0 by convention");
+    }
+
+    [Fact]
+    public void Default_probe_interval_is_reasonable_for_S7_scan_cycle()
+    {
+        // S7 PLCs scan 2-10 ms but comms mailbox typically processed once per scan.
+        // 5 s default probe is lightweight — ~0.001% of comms budget.
+        new S7ProbeOptions().Interval.ShouldBe(TimeSpan.FromSeconds(5));
+    }
+
+    [Fact]
+    public void Tag_definition_defaults_to_writable_with_S7_max_string_length()
+    {
+        var tag = new S7TagDefinition("T", "DB1.DBW0", S7DataType.Int16);
+        tag.Writable.ShouldBeTrue();
+        tag.StringLength.ShouldBe(254, "S7 STRING type max length is 254 chars");
+    }
+
+    [Fact]
+    public void Driver_instance_reports_type_and_id_before_connect()
+    {
+        var opts = new S7DriverOptions { Host = "127.0.0.1" };
+        using var drv = new S7Driver(opts, "s7-test");
+        drv.DriverType.ShouldBe("S7");
+        drv.DriverInstanceId.ShouldBe("s7-test");
+        drv.GetHealth().State.ShouldBe(DriverState.Unknown, "health starts Unknown until InitializeAsync runs");
+    }
+
+    [Fact]
+    public async Task Initialize_against_unreachable_host_transitions_to_Faulted_and_throws()
+    {
+        // Pick an RFC 5737 reserved-for-documentation IP so the connect attempt fails fast
+        // (no DNS mismatch, no accidental traffic to a real PLC).
+        var opts = new S7DriverOptions { Host = "192.0.2.1", Timeout = TimeSpan.FromMilliseconds(250) };
+        using var drv = new S7Driver(opts, "s7-unreach");
+
+        await Should.ThrowAsync<Exception>(async () =>
+            await drv.InitializeAsync("{}", TestContext.Current.CancellationToken));
+
+        var health = drv.GetHealth();
+        health.State.ShouldBe(DriverState.Faulted, "unreachable host must flip the driver to Faulted so operators see it");
+        health.LastError.ShouldNotBeNull();
+    }
+}

tests/ZB.MOM.WW.OtOpcUa.Driver.S7.Tests/ZB.MOM.WW.OtOpcUa.Driver.S7.Tests.csproj

@@ -0,0 +1,31 @@
+<Project Sdk="Microsoft.NET.Sdk">
+
+    <PropertyGroup>
+        <TargetFramework>net10.0</TargetFramework>
+        <Nullable>enable</Nullable>
+        <ImplicitUsings>enable</ImplicitUsings>
+        <IsPackable>false</IsPackable>
+        <IsTestProject>true</IsTestProject>
+        <RootNamespace>ZB.MOM.WW.OtOpcUa.Driver.S7.Tests</RootNamespace>
+    </PropertyGroup>
+
+    <ItemGroup>
+        <PackageReference Include="xunit.v3" Version="1.1.0"/>
+        <PackageReference Include="Shouldly" Version="4.3.0"/>
+        <PackageReference Include="Microsoft.NET.Test.Sdk" Version="17.12.0"/>
+        <PackageReference Include="xunit.runner.visualstudio" Version="3.0.2">
+            <PrivateAssets>all</PrivateAssets>
+            <IncludeAssets>runtime; build; native; contentfiles; analyzers; buildtransitive</IncludeAssets>
+        </PackageReference>
+    </ItemGroup>
+
+    <ItemGroup>
+        <ProjectReference Include="..\..\src\ZB.MOM.WW.OtOpcUa.Driver.S7\ZB.MOM.WW.OtOpcUa.Driver.S7.csproj"/>
+    </ItemGroup>
+
+    <ItemGroup>
+        <NuGetAuditSuppress Include="https://github.com/advisories/GHSA-37gx-xxp4-5rgx"/>
+        <NuGetAuditSuppress Include="https://github.com/advisories/GHSA-w3x6-4m5h-cxqf"/>
+    </ItemGroup>
+
+</Project>