docs(review 007 Unit 2): WS-31 discovery-shadowing note + record corrections

- WS-31: websocket.md §"Data channels for browsers" + ADR-067's landed
  note record that WS-session discovery is the bootstrap set — the
  hook's bootstrap `services/*` registrations overwrite a
  base-registry `services/*` registration on the WS path by design
  (a deployment's custom `services/list` is shadowed on WS sessions
  only).
- ADR-048's landed note: correction + completion — the WS-26
  retention sentence was aspirational at the landed commit (WS-28) and
  is now real; the UP-02 posture's override half is now an explicit
  surface (`with_ws_op_register_acl` / `OpRegisterAcl`), with the
  note that `ChannelsPolicy` could not carry an op ACL.
- ADR-067's landed note: review-007 notes (WS-28 fix + gate, WS-29
  surface, WS-31 record).
- OQ-05 resolution: the retention claim carries the WS-28 correction.
- review-006 UP-02 log + WS-26 paragraph: corrections marking what the
  pre-fix tree did not have, with the landed remediation named.
- review-002 WS-17: the "bounded at 64 sessions" claim corrected —
  the bare-registry semaphore was per-request and bounded nothing;
  `SessionSlots` is the shared-cap surface.
- review-007 status: open for remediation → remediated, with the
  decisions taken (both "implement" options) and the gate names.

Verification: cargo test 454 passed / 0 failed; cargo doc --no-deps
clean (6 pre-existing warnings, identical at baseline).

Review: docs/reviews/007-ws-data-channel-surface-review.md

docs/architecture/decisions/048-websocket-native-session-not-gateway.md

@@ -51,6 +51,22 @@ Accepted
 > `AccessControl::default()` (the SRV-10 permissive-crate-default
 > precedent); deployments gate via a stricter `ChannelsPolicy`
 > passed through the hook (review 006 UP-02 decision).
+>
+> **Review-007 correction + completion (2026-09-05):** the retention
+> sentence above was, at the time of the landed note, aspirational —
+> review 007 WS-28 found the `ConnectionGuard` dropping at the end of
+> its `if let` block (the handle visible for microseconds) and
+> remediation bound it in the channel-0 task's frame; the retention is
+> now real, gated by
+> `live_connections_visible_mid_session_and_drain_after_teardown`.
+> The ACL posture's second half is likewise now real as an explicit
+> surface (the `ChannelsPolicy` object could not carry an op ACL):
+> `HttpAdapter::with_ws_op_register_acl` / the `OpRegisterAcl`
+> request extension set the `op/register` op's `AccessControl` —
+> default unchanged (`AccessControl::default()`, any authenticated
+> peer may announce); a deployment restricting which authenticated
+> peers may announce threads a stricter value. The UP-02 posture
+> stands; only the override surface it named now exists.
 
 ## Status amendment (alkhttp port)
 

docs/architecture/decisions/067-websocket-carries-channels.md

@@ -119,6 +119,18 @@ On upgrade, the handler:
 > (WS-26). Gates: the six e2e scenarios in
 > `tests/ws_upgrade_session.rs` + the `services/list-peers` announced-op
 > discovery gate (alkcall 0.3.1).
+>
+> **Review-007 notes (2026-09-05):** WS-28 fixed the WS-26 retention —
+> the `ConnectionGuard` now lives in the channel-0 task's frame (its
+> drop removes the handle on any teardown path), verified by
+> `live_connections_visible_mid_session_and_drain_after_teardown`.
+> WS-29 implemented the `op/register` override surface the review-006
+> UP-02 posture named: `HttpAdapter::with_ws_op_register_acl` on the
+> built-in surface, the `OpRegisterAcl` request extension on
+> bare-registry routes (default still `AccessControl::default()`).
+> WS-31 recorded: WS-session discovery is the bootstrap set — the
+> hook's bootstrap registrations overwrite a base-registry
+> `services/*` registration on the WS path by design.
 
 Everything ADR-048 says about dispatch — `call.requested` →
 `Dispatcher::dispatch_requested` with `AccessControl::check` gating,