docs: F9/F20-BUG2/F20-BUG3 FIXED (agent v0.31.0, live-validated); queue golden controller-tag follow-up

documentation/audits/live-drive-fixspec-2026-06-14.md

@@ -38,7 +38,16 @@ key fixes live-verified. (F9, F20-BUG2, F20-BUG3 remain for the SUPERVISED agent
 | F5 (catalog) | **FIXED** | app-catalog `main` | uptime-kuma healthy → route 302 (was 404) |
 | F5 (dashboard) | **FIXED** | `803ce50` | funcmap + template render tests |
 | F17 | **FIXED** | `0b9450e` | **marker DB round-trip PASSED live** (escape hatch cleared) |
-| F9, F20-BUG2, F20-BUG3 | DEFERRED | — | SUPERVISED next session (agent/golden) |
+| F9 | **FIXED** | agent `4cd1d02` (+`a2a76e7`), ctrl `0.63.0` | live: agent-restart auto-re-asserted the bind (no manual call); `guest_attached=True`; HDD app deployed onto `/dev/sdb1` |
+| F20-BUG2 | **FIXED** | agent `a2a76e7`, ctrl `0.63.0` | live: confirmed wipe with `/api/disks` `wipe_durable_id` accepted (no binding_mismatch) |
+| F20-BUG3 | **FIXED** | agent `4777f8a` | live on 916GB felhom-usb: 2s client timeout → ~30s mkfs ran to a clean ext4; agent restart mid-format recovered + completed clean |
+
+> **2026-06-14 supervised session (agent v0.31.0 + controller v0.63.0):** F9 / F20-BUG2 / F20-BUG3
+> implemented + deployed + live-validated. Approach was **attach-to-existing** (no re-provision — 9201
+> ran v0.62.0; the golden bakes a stale controller `:0.43.0`, so re-provision would regress it). The
+> golden's stale default controller tag is queued as a separate follow-up (`documentation/backlog/`).
+> Note: the controller has no automated SSD↔HDD data-migration feature (de-privileging removed it) — F9
+> unblocks HDD-app deployment with data on the real HDD, which is what was proven.
 
 Note on F1: the FIXSPEC's "read the cgroup limit" approach proved a **no-op on the demo** — the
 controller container's own cgroup is unlimited (the 2 GB cap is on the LXC ancestor, hidden) and there is