docs(webui): record the tenth review round and fix the round counts

The ninth codex review read the tree the eighth round left behind and raised
two findings. Both were real, and both are fixed on the branch:

- The session declared a liveness ping and never sent one. It held the
  pending-ping table, it answered a pong, and its own option comment said the
  session emits a ping — but no ping was ever written to the socket. An
  admitted session now pings every ten seconds and closes when the pong does
  not come within one interval, so a browser gone behind an open socket is
  noticed and the extension's reconnect ladder takes the session back up.
- The real-browser acceptance failed on a machine without the installed Edge,
  reporting a broken bridge rather than an absent browser. Those cases skip
  with the reason now.

The document also stops carrying two rounds with the same numbers: the two
probe rounds are Probe A and Probe B, and the counts in the header, in the
directory index and in the closing evidence say ten review rounds and
twenty-one acceptance cases.

Author: Fectivnfy112357

packages/webui/docs/browser-use/07-review-findings.md

@@ -1,11 +1,17 @@
 # Review findings — WebUI Browser Use
 
-**Nine review rounds** have been run against this change. Eight of them were run
+**Ten review rounds** have been run against this change. Nine of them were run
 by a `codex` `gpt-6-luna` reviewer reading the source rather than the
 description; the seventh was the final acceptance pass, run against the
 finished tree and against the acceptance criteria the tickets state. This
 document records what each round found and what was done about it.
 
+Two **probe rounds** — a real browser against the built extension, and the
+external path driven against it — are recorded here as well. They are numbered
+`Probe A` and `Probe B` rather than as rounds, because they were findings of
+the implementation, not of a reviewer, and because the two used to carry the
+same numbers as rounds 5 and 6.
+
 Every status below is a claim about the **tree**, and each one names the file
 that carries it. A finding that is still open is marked open; nothing here is
 rounded up.
@@ -321,7 +327,7 @@ tested without a DOM renderer — the client test environment has none.
 - **N4** — The stale claims in this document and in
   `03-extension-admission.md` are corrected.
 
-## Round 5 — found by probing the real browser
+## Probe A — found by probing the real browser
 
 No review round caught this one, because every test of the extension runs
 against its **TypeScript source**. A probe against a real Chrome shows the
@@ -343,7 +349,7 @@ This is the kind of gap that only a real browser finds, and it is the reason
 the programme's evidence says "not yet measured" for the external surfaces
 rather than claiming they work.
 
-## Round 6 — the external path, against a real browser
+## Probe B — the external path, against a real browser
 
 A Playwright spec now drives a **real Chrome with the built extension loaded**
 against a **real bridge server**: `packages/webui/test/browser/browser-bridge-real-browser.spec.ts`.
@@ -609,6 +615,37 @@ eight are closed.
   sealed credential is in the frame that carries it, so the check is about the
   sealing and not about a frame that never arrived.
 
+## Round 10 — the ninth `codex` review, of round 9's fixes
+
+The ninth `codex` review was run against the tree round 9 left behind, with
+the instruction to attack the fixes themselves. It raised two findings; both
+are closed.
+
+| # | Severity | Finding | Status |
+| --- | --- | --- | --- |
+| S1 | MAJOR | **The session declared a liveness ping and never sent one.** It held the pending-ping table, it answered a `pong`, and its own option comment said the session emits a `ping` — but no ping was ever written to the socket. A browser that was gone behind an open socket was therefore never noticed. | `fixed` |
+| S2 | MAJOR | **The real-browser acceptance failed on a machine without the installed Edge.** The Edge cases drive `/usr/bin/microsoft-edge`, so a CI image or a container that does not ship Edge reported a broken bridge rather than an absent browser. | `fixed` |
+
+### What was done
+
+- **S1** — an admitted session now pings the peer every ten seconds and closes
+  the session when the pong does not come within one interval. A WebSocket can
+  stay open long after the browser behind it is gone: the extension's worker is
+  terminated, the profile is closed, the machine sleeps. The socket's close is
+  then the only signal, and it can take a while to arrive; the unanswered ping
+  is the faster signal, and the extension's reconnect ladder picks the session
+  back up. The close is transient at the system level — the peer's state lives
+  in its own storage, so nothing is lost. The extension already answered a
+  `ping` with a `pong`, so nothing on that side changed. The timer is cleared
+  on teardown and does not hold the process open (`session.ts`,
+  `startLiveness`).
+- **S2** — the Edge cases now **skip with the reason** when the installed Edge
+  is not on the machine, rather than failing on a browser that was never there
+  (`real-browser-bridge.ts`, `isBrowserInstalled`). The bundled Chromium is
+  installed by the gate itself, so the Chromium cases always run. The record
+  names the browser each case ran in, and the published matrix names the host,
+  so a skipped case is visible rather than silent.
+
 ## What this document does not establish
 
 - That the external path works end to end against a real Edge or Chrome for

packages/webui/docs/browser-use/README.md

@@ -11,8 +11,8 @@ status line of its own:
 
 Where a document mixes the three, they are separated inline rather than merged into one sentence.
 
-**Read [`07-review-findings.md`](07-review-findings.md) before anything else.** **Nine review rounds**
-have run against this change — eight by a `codex` `gpt-6-luna` reviewer reading the source rather
+**Read [`07-review-findings.md`](07-review-findings.md) before anything else.** **Ten review rounds**
+have run against this change — nine by a `codex` `gpt-6-luna` reviewer reading the source rather
 than the description, and a final acceptance pass against the finished tree. That document records
 every finding and states, for each, whether it is closed or still open. **Every finding is closed**;
 the ones that changed a decision are recorded in the decision record's amendments, and the ones that
@@ -74,7 +74,7 @@ open are listed in [`07-review-findings.md`](07-review-findings.md).
 | [`04-pairing-credentials.md`](04-pairing-credentials.md) | Credential lifecycle, the per-OS secure store on each platform, and the explicit-failure rule when no store exists | You implement pairing, credential storage, revocation or re-enable |
 | [`05-target-identity.md`](05-target-identity.md) | The three-part target identity and its generation, what invalidates a generation, and how a stale reference is rejected | You work with targets, input forwarding, tab selection or run scoping |
 | [`06-runtime-authorization-trace.md`](06-runtime-authorization-trace.md) | Each of the four authorization switches traced to the place it must be enforced, with the current implementation status of each check | You implement an authorization check, or you are auditing whether one exists |
-| [`07-review-findings.md`](07-review-findings.md) | All the findings from the seven review rounds, what was done about each, and what is still open | You want to know what is known to be broken, or what has been corrected |
+| [`07-review-findings.md`](07-review-findings.md) | All the findings from the ten review rounds, what was done about each, and what is still open | You want to know what is known to be broken, or what has been corrected |
 | [`packages/browser-extension/STORE-DISCLOSURES.md`](../../../browser-extension/STORE-DISCLOSURES.md) | The store disclosures for the Chrome Web Store and the Edge Add-ons, and the differences between a packaged build and a development build | You are preparing a store submission, or you need to know what a packaged build is |
 | [`09-acceptance-record.md`](09-acceptance-record.md) | The acceptance matrix: every surface, every case, the browser version, the operating system, the settings and the observed result — written from the records the suites produce | You want to know what was actually run, and where the evidence is thin |