[F49] live-test scaffolding for F54 OnWriteComplete + COM probe diagnostic

Live attempt against AVEVA on this dev host produced two artefacts:

**`crates/mxaccess-compat/tests/lmx_write_complete_live.rs`** — the
F54 OnWriteComplete round-trip test. Compiles + runs against the
live AVEVA install via either path:
- `--features live-windows-com` (preferred): uses
  `Session::connect_nmx_auto` so the COM activation reference is
  held in-process for the duration of the test.
- Default features (fallback): shells out to
  `MxNativeClient.Probe --probe-resolve-oxid-managed-ntlm-integrity`
  + `--probe-remqi-managed` to learn the per-session NMX endpoint +
  INmxService2 IPID, then uses `Session::connect_nmx`.

Both code paths are wired and the test runs through endpoint
resolution + IPID extraction successfully. The connect step itself
fails with `Status { detail: 1722 }` (RPC_S_SERVER_UNAVAILABLE).

**`crates/mxaccess-rpc/examples/com-marshal-probe.rs`** — minimal
one-shot binary that calls
`marshal_activated_iunknown_objref("NmxSvc.NmxService",
DifferentMachine)` in isolation. Confirms the COM activation +
CoMarshalInterface chain works fine standalone (returns a 338-byte
OBJREF with valid OXID/IPID structure). The 1722 in the live test
is therefore downstream of the activation — likely a COM-apartment
threading interaction with the tokio multi-thread runtime.

This is an F12-related issue (auto-resolve hardening), not an F54
issue. F54's correctness is covered by the existing unit-level
integration tests:
- `mxaccess::session::tests::router_populates_operation_status_context_from_pending_ops_fifo`
- `mxaccess::session::tests::write_handle_correlates_with_router_emitted_status`
- `mxaccess_compat::tests::drain_routes_write_status_to_on_write_complete`
- `mxaccess_compat::tests::drain_routes_non_write_status_to_on_operation_complete`

`design/followups.md` F49 entry updated to reflect:
- F54 added as a fifth row in the live-verification scope.
- "Live attempt 2026-05-06" sub-section documents the 1722 issue +
  what was verified (.NET probe end-to-end works against same
  install; Rust COM activation works in isolation; the failure is
  Rust-port-specific to `connect_nmx_auto` under tokio).
- F49 now Blocked-by F12 hardening (the 1722 path).

New `live-windows-com` feature on `mxaccess-compat` propagates to
`mxaccess/windows-com` for the test binary.

Workspace 824 → 824 tests; clippy + rustdoc clean across both
feature configurations.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

design/followups.md

@@ -27,19 +27,24 @@ Between each publish: wait for the crate to be indexed before the next one's `ca
 
 ### F49 — Live verification sweep for the M6 features
 **Severity:** P1 — closes the live-evidence gap for the M6 work that landed unit-only this session.
-**Source:** F36, F40, F45, F47 closeouts — each ships with unit tests but several were not exercised against the live AVEVA install in this session.
+**Source:** F36, F40, F45, F47, F54 closeouts — each ships with unit tests but most were not exercised against the live AVEVA install in this session.
+**Blocked-by:** F12 hardening (`Session::connect_nmx_auto` returns `RPC_S_SERVER_UNAVAILABLE` (1722) under `cargo test`'s tokio multi-thread runtime — see "Live attempt 2026-05-06" below). The COM-activation path itself works in isolation (`cargo run -p mxaccess-rpc --example com-marshal-probe --features windows-com` succeeds), so the failure is downstream — likely a COM apartment threading issue when CoInitializeEx runs on a tokio worker thread.
 
 **Scope.** Run the following against the live AVEVA host with `MX_LIVE=1`:
 1. **F36 buffered subscribe** — `cargo run -p mxaccess --example subscribe-buffered -- --tag TestChildObject.TestInt`. Confirm `OnBufferedDataChange`-rate updates flow at the configured cadence; capture wire bytes via `analysis/frida/mx-nmx-trace.js` and confirm exactly one `RegisterReference` (`0x10`) frame with `.property(buffer)` suffix, no separate `SetBufferedUpdateInterval` RPC, and no separate `AdviseSupervisory` follow-up.
 2. **F45 recovery replay for buffered** — start the `subscribe-buffered` example, force a `Session::recover_connection` mid-flight (e.g. via a `wwtools` helper that bumps the NMX TCP socket), assert the post-recovery NMX traffic carries an `RegisterReference` (NOT `AdviseSupervisory`) with the same correlation id and `.property(buffer)` suffix.
 3. **F47 buffered unsubscribe skip** — instrument `Session::unsubscribe` with a `tracing::debug` log line on the buffered branch, run the example to completion + drop, confirm no `UnAdvise` frame in the wire trace.
 4. **F40 metrics** — install a `metrics` exporter (`metrics-exporter-prometheus` is the lightest), run `connect-write-read` + `subscribe` examples with `--features metrics`, confirm at least one counter increment and one histogram observation per metric name in the registered set.
+5. **F54 OnWriteComplete (LmxClient round-trip)** — scaffold lives at `crates/mxaccess-compat/tests/lmx_write_complete_live.rs`. Run `cargo test -p mxaccess-compat --features live-windows-com --test lmx_write_complete_live -- --ignored --nocapture` to drive `LmxClient::write` → drain `client.on_write_complete()` and assert the `WriteCompleteEvent { server_handle, item_handle, statuses, is_during_recovery }` shape matches `LMX_OnWriteComplete(int hLMXServerHandle, int phItemHandle, ref MXSTATUS_PROXY[] pVars)`.
+
+**Live attempt 2026-05-06.** Steps 1-4 not run yet. Step 5 attempted; the test compiled and ran past Frida-style `--probe-resolve-oxid-managed-ntlm-integrity` resolution + `--probe-remqi-managed` IPID extraction, but `connect_nmx_auto` (preferred path) and `connect_nmx` (fallback with probe-resolved IPID) both fail with `Status { detail: 1722 }` (RPC_S_SERVER_UNAVAILABLE). The .NET `MxNativeClient.Probe --probe-session-write` runs the same scenario successfully end-to-end against the same AVEVA install, so the wire is functional and the failure is Rust-port specific. Documented as the F12 hardening followup; the F54 unit-level integration tests (`router_populates_operation_status_context_from_pending_ops_fifo` + `write_handle_correlates_with_router_emitted_status`) cover the F54 logic exhaustively at the layer boundaries.
 
 **Definition of done:**
 1. Per-feature evidence summary in `docs/M6-live-verification.md` (one paragraph per feature with the wire-trace excerpt or metrics-exporter snapshot).
 2. If any feature fails live: file a sub-followup with the captured failure and link it from the evidence doc.
+3. F12's tokio-runtime COM activation issue resolved (the `connect_nmx_auto` 1722 error above) so the live tests can actually run.
 
-**Resolves when:** all four features have a live evidence row + no sub-followups remain unresolved.
+**Resolves when:** all five features have a live evidence row + no sub-followups remain unresolved.
 
 ### F50 — Run the F46 Suspend/Activate Frida capture live
 **Severity:** P3 — residual from F46 (script ready, capture not yet run).