[ab-check] installer ran hydrogen.exe --check-and-upgrade-ab (a helper run, not a start)
21:20:20 [ad-install] no bundled ab installer staged (thin build) -> resolving the latest ab from the wiki
21:20:20 [ad-install] latest ab = 2.1.168 via version.json (wiki v2) (https://wiki.adom.inc/download/adom/adom-bridge/2.1.168/Adom.Bridge_2.1.168_windows_x64-setup.exe)
21:20:20 [ad-install] ab not detected -> installing
21:20:21 [ad-install] downloaded 21586680 bytes
21:20:21 [ad-install] sha256 verified against the published release
21:20:31 [ad-install] ab installer exited 0
21:20:31 [ad-install] wrote ab-installed-by-ah.flag (ah owns this ab install; uninstall may cascade)
21:20:31 [ad-install] ab 2.1.168 installed from the wiki

=== Hydrogen Desktop started at 2026-10-07 21:20:31 ===
[app] BUILD_STAMP=fcd70f336be29b0a91cb12c1300816a9c0abb3af-dirty-1791216615
21:20:31.687 [ports] No ports.json found: using defaults: code_server=7380 proxy=47083 control=47084 cdp=47085
21:20:32.493 [app] Starting Tauri builder... (hd-rust-20260519-bp01) pid=9280 exe=C:\Users\garyc\AppData\Local\Adom Hydrogen\hydrogen.exe
21:20:32.494 [app] instance pid=9280 args=[]
21:20:32.495 [ports] PortConfig: proxy=47083, control=47084, cdp=47085, code_server=7380
21:20:32.499 [discover] Discovery server on http://127.0.0.1:47080/discover
21:20:32.503 [net] WSL networkingMode=mirrored detected, host services bound loopback-only (no LAN, no firewall prompt).
21:20:32.504 [direct-api] DEPRECATED shim on http://127.0.0.1:8770: use the control API's /direct/* routes; this listener is removed next release (see docs/PORTS-CONSOLIDATION.md)
21:20:32.530 [app] Tauri app built successfully, entering run loop
21:20:32.533 [deep-link] hydrogen:// scheme registered for this user
21:20:32.533 [wake-heal] boot: Hydrogen started: recovery watchers armed (Win32 power notifications)
21:20:32.534 [main-watchdog] armed (10s main-thread ping, 20s stall threshold, self-minidump on first stall)
21:20:32.534 [app] previous run ended UNCLEAN (no clean-shutdown marker) — will heal the VS Code layout after restore
21:20:32.534 [app] Setup callback entered
21:20:32.534 [app] Active workspace runtime: WslDistro
21:20:32.536 [wake-heal] suspend/resume notifications registered
21:20:32.592 [app] Tray icon installed
21:20:32.593 [dev] No build/ directory found — dev file server not started. Run 'pnpm build' first.
21:20:32.594 [app] window size: 1580x1008 (95% of the primary monitor's work area, floor 1400x900)
21:20:32.594 [wsl-trigger] backend poller armed (C:\Users\garyc\AppData\Roaming\hydrogen-desktop\setup-trigger-wsl.json)
21:20:32.599 [ad-supervisor] HD liveness beacon listening on 127.0.0.1:9001 (AB embedded probe)
21:20:32.700 [autostart] enabled: Run key written as 'Adom Hydrogen' (legacy value cleared) + StartupApproved disabled-flags cleared
21:20:32.895 [app] 'hd' window created with URL interception handlers
21:20:33.044 [first-run] native glow backdrop up (2496x1664, frost blurred at 1/4)
21:20:33.057 [on-navigation] ALL: http://tauri.localhost/
21:20:33.172 [telemetry] first_launch: HTTP 200
21:20:33.271 [taskbar] added a taskbar button for the onboarding card (hwnd=67300); it is owned by the backdrop, so the shell skips it otherwise
21:20:33.271 [glass] click-through armed on the onboarding card
21:20:33.274 [first-run] onboarding card foregrounded once at open (the only unsolicited raise; auth handbacks are the others)
21:20:33.275 [first-run] onboarding card open (main window hidden until dismissed)
21:20:33.275 [window-state] first run: ignoring any saved geometry, centering (centered default 1580x1008)
21:20:33.276 [icon] ICON CODE REACHED - about to load icon.png (3815 bytes)
21:20:33.276 [icon] from_bytes result: OK
21:20:33.276 [icon] set_icon SUCCESS
21:20:33.277 [chrome] native window base brush set to #1f2125 (prev=0x0) - kills the maximized edge seam and the launch white flash; the chrome DOCUMENT stays transparent
21:20:33.277 [window] nc clamp armed (maximized client rect clamped to the monitor work area)
21:20:33.277 [script-dialog] default script dialogs DISABLED (beforeunload 'Leave site?' suppression)
21:20:33.277 [script-dialog] Registered ScriptDialogOpening auto-accept handler
21:20:33.277 [zoom] AcceleratorKeyPressed registered, Ctrl +/-/0 now work from inside the panels
21:20:33.278 [frame-nav] Registered FrameNavigationStarting handler
21:20:33.278 [raw-new-window] Registered NewWindowRequested handler
21:20:33.278 [ctx-menu] Registered ContextMenuRequested handler
21:20:33.278 [script] Registered target=_blank click interceptor for all frames
21:20:33.278 [script] Registered the non-page-scheme link guard for all frames
21:20:33.278 [perm] Registered PermissionRequested auto-grant handler
21:20:33.279 [window-state] periodic snapshot started
21:20:33.280 [proxy] Listening on http://127.0.0.1:47083
21:20:33.281 [discovery] responder on http://127.0.0.1:47080
21:20:33.281 [control] AI control API on http://127.0.0.1:47084 (build 2026-05-22T13)
21:20:36.699 [app] WebView2 processes at above-normal priority: 7 (pass at +5s, adom/hydrogen#75)
21:20:37.603 [wsl-startup] state=never-run → auto-running setup cascade (force=false)
21:20:37.603 [wsl-startup] Adom sign-in in flight (no session token yet) — NOT foregrounding HD; cascade builds in background, HD comes forward on auth completion
21:20:39.152 [run_all_wsl] no workspace distro yet: prefetching the golden image in the background
21:20:39.155 [taskbar] first paint on hwnd=67300 (0/1000) state_ok=true value_ok=true
21:20:39.180 [wsl-tarball] fetching https://github.com/adom-inc/hd-wsl2-image/releases/download/v31-full/adom-golden-v31-full.tar.gz (resume from byte 0)
21:20:39.687 [wsl-tarball prefetch] ↓ downloading golden image, 775 MB total…
21:20:40.501 [run_all_wsl] resume: keep done steps, run from first non-done (distro_real=false)
21:20:40.502 [run_all_wsl] start: force=false steps=8 state_entries=1
21:20:40.502 [cascade] setup OWNS the editor (opportunistic healers stand down)
21:20:40.627 [wsl-startup] code-server did not answer over HTTP within 8033 ms: the launch tasks run now, failures are observe-only for the boot window
21:20:40.664 [wsl-startup] distro not registered yet — code-server re-ensure deferred to cascade
21:20:42.744 [wsl-tarball prefetch] ↓ golden image: 130/775 MB (17%) @ 356.0 Mbps, ETA ~15s
21:20:43.281 [heartbeat] alive 10s
21:20:45.795 [wsl-tarball prefetch] ↓ golden image: 220/775 MB (28%) @ 302.2 Mbps, ETA ~15s
21:20:46.235 [ad-supervisor] spawned AB detached (breakaway) pid 17904
21:20:46.235 [ad-supervisor] spawned AB pid=17904 before the relay check (token=absent)
21:20:46.374 [ad-relay] relay :8766 not answering /health, restarting adom-relay.service once
21:20:48.801 [wsl-tarball prefetch] ↓ golden image: 350/775 MB (45%) @ 322.2 Mbps, ETA ~11s
21:20:51.064 [run_all_wsl] install-approval not available (AB direct API not reachable (no live discovery file + fixed range silent, likely HNS-reserved)), continuing with per-verb prompts
21:20:51.064 [run_all_wsl] RUN 'ensure-workspace' (status was Pending)
21:20:51.200 [telemetry] setup_step ensure-workspace start: HTTP 200
21:20:51.909 [wsl-tarball prefetch] ↓ golden image: 449/775 MB (58%) @ 308.1 Mbps, ETA ~8s
21:20:52.704 Update check: on latest (1.0.545 == 1.0.545)
21:20:52.892 [updater] fleet ping HTTP 200 (user '')
21:20:53.281 [heartbeat] alive 20s
21:20:53.294 [window-state] snapshot saved: 62,40 2371x1512 max=false
21:20:54.913 [wsl-tarball prefetch] ↓ golden image: 589/775 MB (76%) @ 324.6 Mbps, ETA ~4s
21:20:56.715 [app] WebView2 processes at above-normal priority: 13 (pass at +20s, adom/hydrogen#75)
21:20:58.021 [wsl-tarball prefetch] ↓ golden image: 715/775 MB (92%) @ 327.2 Mbps, ETA ~1s
21:20:59.295 [wsl-tarball prefetch] ✓ golden image downloaded, 775 MB in 19s (331.7 Mbps avg)
21:20:59.295 [wsl-tarball] downloaded 813007191 bytes to "C:\\Users\\garyc\\AppData\\Local\\Adom\\HydrogenDesktop\\cache\\adom-golden.tar.gz"
21:20:59.296 [wsl-tarball prefetch] image ready at "C:\\Users\\garyc\\AppData\\Local\\Adom\\HydrogenDesktop\\cache\\adom-golden.tar.gz"
21:20:59.956 [ad-relay] relay STILL down after restart (ws:8765 http:8766); status: workspace is still booting, try again in a moment
21:20:59.956 [ad-supervisor] relay :8765 not answering at supervisor start, ab will show red until it recovers
21:20:59.959 [ad-supervisor] AB already running - adopting into embedded mode (POST /embedded/enter)
21:20:59.967 [ad-supervisor] embedded/enter ok: {"_hint":"ab is now embedded as runtime-adopt. When ah dies, ab will revert to standalone (NOT exit). If ah wants ab to exit on ah-death, spawn ab with `--embedded` instead.","embedded":true,"entered_via":"launch-flag","ok":true,"owner":"hydrogen","schemaVersion":1,"status":"ok","success":true,"was_embedded":true}
21:21:02.238 [wsl-tarball] SHA256 OK (813007191 bytes)
21:21:03.282 [heartbeat] alive 30s
21:21:04.989 [ad-supervisor] brand fonts: 0 installed now, 3 present, 0 failed
21:21:13.283 [heartbeat] alive 40s
21:21:23.284 [heartbeat] alive 50s
21:21:25.776 [wsl] ["-d", "Adom-Workspace", "--exec", "/bin/true"] failed on the management plane (wsl ["-d", "Adom-Workspace", "--exec", "/bin/true"] exited with Some(-); retry 1/3 in 2 s (53 s after boot: observe only)
21:21:28.271 [wsl] ["-d", "Adom-Workspace", "--exec", "/bin/true"] failed on the management plane (wsl ["-d", "Adom-Workspace", "--exec", "/bin/true"] exited with Some(-); retry 2/3 in 5 s (53 s after boot: observe only)
21:21:33.285 [heartbeat] alive 60s
21:21:33.728 [wsl] ["-d", "Adom-Workspace", "--exec", "/bin/true"] failed on the management plane (wsl ["-d", "Adom-Workspace", "--exec", "/bin/true"] exited with Some(-); retry 3/3 in 15 s (53 s after boot: observe only)
21:21:43.286 [heartbeat] alive 70s
21:21:49.295 [wsl-gate] 'wsl -d (management plane)' timed out: holding every workspace command for 60s instead of stacking more (trip #1)
21:21:53.287 [heartbeat] alive 80s
21:21:56.726 [app] WebView2 processes at above-normal priority: 13 (pass at +60s, adom/hydrogen#75)
21:22:03.288 [heartbeat] alive 90s
21:22:07.779 [run_all_wsl] DONE 'ensure-workspace' in 76.7s (failed)
21:22:07.791 [run_all_wsl] HALT: step 'ensure-workspace' failed after 1 attempts; not advancing to later steps
21:22:07.793 [run_all_wsl] persisted setup-state.txt = in-progress (GET /setup/state matches the cascade)
21:22:07.793 [run_all_wsl] TIMING total 76.7s over 1 steps; slowest: ensure-workspace 76.7s
21:22:07.795 [cascade] setup released the editor (opportunistic healers re-armed)
21:22:07.799 [run_all_wsl] install-approval window closed, normal prompting restored
21:22:07.800 [taskbar] cascade ended with a failure, leaving the red progress bar up
21:22:07.800 [wsl-startup] auto-run run_all_wsl finished: Ok("{\"completed\":0,\"failed\":1,\"failures\":[{\"error\":\"=== Ensure workspace runtime (WSL2 Ubuntu) ===\\n=== Ensure WSL2 subsystem is responsive (auto-recovers if wedged) ===\\n  $ wsl --status\\n    | Default Distribution: Ubuntu-24.04\\n    | Default Version: 2\\n  ✓ `wsl --status` ok (0s)\\n  ✓ WSL2 subsystem is responsive\\nChecking WSL2 prerequisites (wsl --status)...\\n  ✓ WSL2 is installed and ready\\n=== Ensure WSL2 subsystem is responsive (auto-recovers if wedged) ===\\n  $ wsl --status\\n    | Default Distribution: Ubuntu-24.04\\n    | Default Version: 2\\n  ✓ `wsl --status` ok (0s)\\n  ✓ WSL2 subsystem is responsive\\nDistro 'Adom-Workspace' not registered, importing the golden image.\\n  → fetching the pre-baked workspace image (≈585 MB)…\\n  ✓ golden image already downloaded (775 MB cached, SHA256 OK), skipping the download\\n  ✓ integrity already verified during the cache check (SHA256 OK)\\n  → wsl --import Adom-Workspace (this can take a moment)...\\n    … extracting rootfs into the workspace VHD, 12 MB written, 3s elapsed…\\n    … extracting rootfs into the workspace VHD, 44 MB written, 6s elapsed…\\n    … extracting rootfs into the workspace VHD, 204 MB written, 9s elapsed…\\n    … extracting rootfs into the workspace VHD, 780 MB written, 12s elapsed…\\n    … extracting rootfs into the workspace VHD, 1452 MB written, 15s elapsed…\\n    … extracting rootfs into the workspace VHD, 2188 MB written, 18s elapsed…\\n    … extracting rootfs into the workspace VHD, 2476 MB written, 21s elapsed…\\n  ✓ imported the v31-full image (775 MB SHA-verified tarball; this is FIXED every run). Workspace VHD: ~2476 MB on disk. This one VARIES run-to-run on purpose: ext4.vhdx is a live, dynamically-growing volume (ext4 overhead + WSL allocation + boot/disk activity), so it's disk usage, not a fixed image total. (21s)\\n  ⚠ import attempt 1/4 failed: import reported success but the distro is not exec-able\\n  → clearing any stale / half-registered 'Adom-Workspace' entry before retrying (scoped wsl --unregister; Docker-safe; no-op if none)…\\n  ✓ cleared a stale workspace entry (or there was none), the next import has a clean name\\n  … probing the WSL subsystem (`wsl --status`), will retry import the moment it's responsive…\\n  … cleared a partial import directory left by the previous attempt\\n    … extracting rootfs into the workspace VHD, 44 MB written, 3s elapsed…\\n    … extracting rootfs into the workspace VHD, 204 MB written, 6s elapsed…\\n    … extracting rootfs into the workspace VHD, 844 MB written, 9s elapsed…\\n    … extracting rootfs into the workspace VHD, 1452 MB written, 12s elapsed…\\n    … extracting rootfs into the workspace VHD, 2444 MB written, 15s elapsed…\\n  ✓ imported the v31-full image (775 MB SHA-verified tarball; this is FIXED every run). Workspace VHD: ~2476 MB on disk. This one VARIES run-to-run on purpose: ext4.vhdx is a live, dynamically-growing volume (ext4 overhead + WSL allocation + boot/disk activity), so it's disk usage, not a fixed image total. (17s)\\n  ⚠ import attempt 2/4 failed: import reported success but the distro is not exec-able\\n  … probing the WSL subsystem (`wsl --status`), will retry import the moment it's responsive…\\n  ✓ imported the v31-full image (775 MB SHA-verified tarball; this is FIXED every run). Workspace VHD: ~2476 MB on disk. This one VARIES run-to-run on purpose: ext4.vhdx is a live, dynamically-growing volume (ext4 overhead + WSL allocation + boot/disk activity), so it's disk usage, not a fixed image total. (0s)\\n  ⚠ import attempt 3/4 failed: wsl [\\\"--import\\\", \\\"Adom-Workspace\\\", \\\"C:\\\\\\\\Users\\\\\\\\garyc\\\\\\\\AppData\\\\\\\\Local\\\\\\\\Adom\\\\\\\\HydrogenDesktop\\\\\\\\wsl\\\\\\\\Adom-Workspace\\\", \\\"C:\\\\\\\\Users\\\\\\\\garyc\\\\\\\\AppData\\\\\\\\Local\\\\\\\\Adom\\\\\\\\HydrogenDesktop\\\\\\\\cache\\\\\\\\adom-golden.tar.gz\\\", \\\"--version\\\", \\\"2\\\"] exited with Some(-1): A distribution with the supplied name already exists. Use --name to choose a different name.\\n  ⚠ import still failing, terminating Adom-Workspace before the last attempt (other distros untouched)…\\n  … probing the WSL subsystem (`wsl --status`), will retry import the moment it's responsive…\\n  ✓ imported the v31-full image (775 MB SHA-verified tarball; this is FIXED every run). Workspace VHD: ~2476 MB on disk. This one VARIES run-to-run on purpose: ext4.vhdx is a live, dynamically-growing volume (ext4 overhead + WSL allocation + boot/disk activity), so it's disk usage, not a fixed image total. (0s)\\n  ⚠ import attempt 4/4 failed: wsl [\\\"--import\\\", \\\"Adom-Workspace\\\", \\\"C:\\\\\\\\Users\\\\\\\\garyc\\\\\\\\AppData\\\\\\\\Local\\\\\\\\Adom\\\\\\\\HydrogenDesktop\\\\\\\\wsl\\\\\\\\Adom-Workspace\\\", \\\"C:\\\\\\\\Users\\\\\\\\garyc\\\\\\\\AppData\\\\\\\\Local\\\\\\\\Adom\\\\\\\\HydrogenDesktop\\\\\\\\cache\\\\\\\\adom-golden.tar.gz\\\", \\\"--version\\\", \\\"2\\\"] exited with Some(-1): A distribution with the supplied name already exists. Use --name to choose a different name.\\n  ✗ import failed: wsl --import failed after 4 attempts: wsl [\\\"--import\\\", \\\"Adom-Workspace\\\", \\\"C:\\\\\\\\Users\\\\\\\\garyc\\\\\\\\AppData\\\\\\\\Local\\\\\\\\Adom\\\\\\\\HydrogenDesktop\\\\\\\\wsl\\\\\\\\Adom-Workspace\\\", \\\"C:\\\\\\\\Users\\\\\\\\garyc\\\\\\\\AppData\\\\\\\\Local\\\\\\\\Adom\\\\\\\\HydrogenDesktop\\\\\\\\cache\\\\\\\\adom-golden.tar.gz\\\", \\\"--version\\\", \\\"2\\\"] exited with Some(-1): A distribution with the supplied name already exists. Use --name to choose a different name.\\r\\nError code: Wsl/Service/RegisterDistro/ERROR_ALREADY_EXISTS\\r\\nIf this error was unexpected, please consider searching for existing issues or filing a new issue at https://github.com/microsoft/WSL/issues.\\n  ✗ workspace setup attempt failed (will retry if attempts remain): wsl --import failed after 4 attempts: wsl [\\\"--import\\\", \\\"Adom-Workspace\\\", \\\"C:\\\\\\\\Users\\\\\\\\garyc\\\\\\\\AppData\\\\\\\\Local\\\\\\\\Adom\\\\\\\\HydrogenDesktop\\\\\\\\wsl\\\\\\\\Adom-Workspace\\\", \\\"C:\\\\\\\\Users\\\\\\\\garyc\\\\\\\\AppData\\\\\\\\Local\\\\\\\\Adom\\\\\\\\HydrogenDesktop\\\\\\\\cache\\\\\\\\adom-golden.tar.gz\\\", \\\"--version\\\", \\\"2\\\"] exited with Some(-1): A distribution with the supplied name already exists. Use --name to choose a different name.\\r\\nError code: Wsl/Service/RegisterDistro/ERROR_ALREADY_EXISTS\\r\\nIf this error was unexpected, please consider searching for existing issues or filing a new issue at https://github.com/microsoft/WSL/issues.\",\"step\":\"ensure-workspace\"}],\"ok\":false,\"timings\":[{\"id\":\"ensure-workspace\",\"ms\":76714,\"ok\":false}],\"total\":8,\"total_ms\":76729}")
21:22:07.800 [wsl-startup] post-cascade: started code-server update poller
21:22:08.104 [telemetry] setup_step ensure-workspace fail: HTTP 200
21:22:13.288 [heartbeat] alive 100s
21:22:23.289 [heartbeat] alive 110s
21:22:33.290 [heartbeat] alive 120s
21:22:43.290 [heartbeat] alive 130s
21:22:53.294 [heartbeat] alive 140s
21:23:03.296 [heartbeat] alive 150s
21:23:03.992 [ad-supervisor] AB health probe failed (1/3 consecutive) - debouncing, no respawn yet
21:23:11.502 [wsl] ["-d", "Adom-Workspace", "-u", "adom", "--", "setsid", "bash", "-lc", "/usr/lib/code-server/bin/code-server --version 2>/dev/null | grep -oE '[0-9]+(\\.[0-9]+)+' | head -1"] failed on the management plane (wsl ["-d", "Adom-Workspace", "-u", "adom", "--", "setsid", "bash", "-l); retry 1/3 in 2 s
21:23:13.296 [heartbeat] alive 160s
21:23:14.045 [wsl] ["-d", "Adom-Workspace", "-u", "adom", "--", "setsid", "bash", "-lc", "/usr/lib/code-server/bin/code-server --version 2>/dev/null | grep -oE '[0-9]+(\\.[0-9]+)+' | head -1"] failed on the management plane (wsl ["-d", "Adom-Workspace", "-u", "adom", "--", "setsid", "bash", "-l); retry 2/3 in 5 s
21:23:19.411 [wsl] ["-d", "Adom-Workspace", "-u", "adom", "--", "setsid", "bash", "-lc", "/usr/lib/code-server/bin/code-server --version 2>/dev/null | grep -oE '[0-9]+(\\.[0-9]+)+' | head -1"] failed on the management plane (wsl ["-d", "Adom-Workspace", "-u", "adom", "--", "setsid", "bash", "-l); retry 3/3 in 15 s
21:23:23.298 [heartbeat] alive 170s
21:23:33.299 [heartbeat] alive 180s
21:23:34.779 [wsl-gate] 'wsl -d (management plane)' timed out: holding every workspace command for 60s instead of stacking more (trip #2)
21:23:34.830 [extensions] could not read the installed set (workspace not ready?); skipping the check
21:23:43.300 [heartbeat] alive 190s
21:23:53.300 [heartbeat] alive 200s
21:23:58.389 [first-run] focus_signin_window: we have not opened a sign-in for this intent, so there is nothing to raise
21:23:58.502 [auth] Auth intent created, URL: https://hydrogen.adom.inc/auth/intent?token=[redacted]
21:23:58.502 [auth-ttl] adom intent created claimed_ttl=904s expires_at=2026-10-08T04:39:03Z
21:23:58.503 [auth] backend poll armed for 934s (intent ttl 904s)
21:23:58.508 [browser] open-in-profile: https://hydrogen.adom.inc/auth/intent?token=[redacted] in default:Default fresh=true
21:23:58.508 [caption] auth narration suppressed (assisted sign-in OFF): Opening the Adom sign-in in your default browser (Default)…
21:23:58.829 [browser] raised the opened window hwnd=786546 for the user -> true
21:23:58.830 [approve] assisted sign-in OFF, Adom auto-approve loop NOT armed; the user clicks Approve themselves
21:23:58.831 [browser] recorded consent window hwnd=786546 (direct open), so it can be raised and closed later
21:24:02.896 [auth] Intent authenticated, session token saved (+ replay copy)
21:24:02.896 [auth] Bringing HD to the foreground after Adom auth completion…
21:24:02.978 [auth] workspace not running, login token propagates on next start
21:24:03.095 [updater] signed-in user changed, checking now so the fleet ping carries who this is