← § Help

Keychain Troubleshooting

Quick Answer

"Keychain unavailable" means the Hollowmark daemon could not read its saved sign-in credential from its local, protected credential store. Sync pauses until it's resolved — your account and your collection/match data are not affected. Quit and relaunch the daemon first; if that doesn't clear it, on a Mac, reinstalling the daemon from Settings → Danger Zone → Uninstall Daemon — with "Also wipe my local config + cached data (irreversible)" ticked — resolves most remaining cases. On Windows, report it instead (see "Still stuck?"). One narrower case (a failed one-time credential migration right after an upgrade) can't be cleared from the app today; if that's what you're seeing, skip reinstalling and report it instead.

What "Keychain unavailable" means

As of Hollowmark v0.4.3, the daemon keeps its sign-in credential in a small, protected file inside its own private config folder on your machine — not in the macOS login Keychain or Windows Credential Manager as the default store. (On Windows, Credential Manager is still the store — see below.) The macOS system Keychain is now consulted only once, as a read-only migration source, the first time the daemon runs after upgrading from an older version that did store it there; after that one-time migration, the daemon never reads the system Keychain again. The service name com.hollowmark.daemon you may see referenced (for example in Keychain Access) is that legacy/ migration-source entry, not the daemon's day-to-day credential.

When the daemon can't read that credential — the file, or on Windows the Credential Manager entry — it can't confirm you're signed in, so it pauses syncing rather than guessing. This shows up in the app as a red Keychain unavailable indicator on the Setup page. Nothing about your Hollowmark account, your collection, or your match history is lost or at risk — this is strictly about the daemon's local access to its own stored credential on this machine.

macOS causes

The daemon's local credential file has become unreadable or corrupted. This is the dominant cause on any machine running v0.4.3 or later. The daemon runs as a background process (a launchd agent) with no window and can't show you a dialog, so if something changes the file's permissions or ownership — a backup tool, a disk-repair utility, or another process that touched the daemon's config folder with elevated privileges — the daemon silently can't read the file it needs and reports "Keychain unavailable" with no visible prompt.

One-time exception — right after upgrading. If you just updated Hollowmark from an older version, the very first launch does a one-time, read-only check of the old macOS Keychain entry to migrate it into the new file. If that read is blocked, you can see this same error on that first post-upgrade launch specifically — and, same as above, it happens silently: a background launchd process gets no dialog to respond to, it just fails the read.

For most affected installs, a later relaunch succeeds once whatever blocked that one read has cleared. But if your old Keychain entry is the kind a background process can never read — no dialog ever appears for it to answer — the daemon retries the same migration attempt, and fails the same way, on every subsequent launch. If that's your situation, there is currently no in-app step that clears it, including reinstalling the daemon (even with the local-config wipe described below ticked): reinstalling replaces the daemon's own credential file, but it does not touch the old system Keychain entry this specific case depends on, so the same blocked read happens again on the next launch. If "Try Again" (below) doesn't clear it and this description matches what you're seeing, skip reinstalling and go straight to "Still stuck?" — use the in-app bug report or Discord #bugs so we can look at it directly. We're tracking the underlying product gap separately.

What does not cause this (despite how it may look): a locked login Keychain, a denied Keychain access prompt, or a login-Keychain password out of sync with your Mac password. Those all describe the old storage mechanism, which the daemon stopped using in v0.4.3. Opening Keychain Access and unlocking or fixing the login keychain does not touch the file the daemon actually reads, so it will not fix a "Keychain unavailable" error on a current install.

Not to be confused with: the one-time admin-password dialog you may see when granting the collection helper permission to read your MTG Arena collection (Settings → Grant Access…). That's a separate, unrelated consent flow for a different component — it has nothing to do with the daemon's credential store, and resolving one does not affect the other.

Fixing it on macOS

  1. Quit and relaunch the Hollowmark daemon (menu bar icon > Quit, then relaunch from Applications). A fresh process gets a fresh attempt at reading the credential file, and this alone clears transient cases.

  2. If it's still showing "Keychain unavailable," click the menu bar icon and choose "Try Again." This is the daemon's actual retry control for this state — there is currently no equivalent action on the in-app Setup page itself, so look for it in the menu bar, not in Hollowmark's Setup screen.

  3. If "Try Again" doesn't clear it, reinstall the daemon — and tick the wipe checkbox. This is the expected fix when the credential file itself is damaged or has bad permissions. Go to Settings → Danger Zone → Uninstall Daemon (see Daemon Uninstall); in the confirmation dialog, tick "Also wipe my local config + cached data (irreversible)" before clicking confirm. That checkbox is what actually deletes the corrupted credential file — it's unchecked by default, and leaving it unchecked, or just reinstalling the app itself, leaves the old file untouched and you'll hit the same error again. With the box ticked, you get a fresh config folder with a fresh, readable credential store. Reinstall from the Installation Guide and sign in again. Your Hollowmark account and all your collection/match data are unaffected; none of that lives on your machine.

    Exception: if your symptom matches the one-time upgrade-migration case described above, ticking the box won't help — that case doesn't involve the credential file this checkbox clears. Skip to "Still stuck?" instead.

Windows causes

The Windows Credential Manager service is unavailable or disabled. Windows Credential Manager runs as a background service; if it has been disabled (by a group policy, security tool, or manual change), no application — including the Hollowmark daemon — can store or retrieve a credential through it.

Fixing it on Windows

  1. Quit and relaunch the Hollowmark daemon (right-click the system tray icon > Exit, then relaunch from the Start Menu). This is the fastest fix if Credential Manager was only temporarily unavailable (for example, immediately after a Windows update).

  2. Confirm Credential Manager is enabled. Open Control Panel > User Accounts > Credential Manager and confirm the window opens normally. If it fails to open or shows an error, your machine's Credential Manager service has been disabled — this is usually done by a corporate IT policy or a third-party security tool. Contact your IT administrator if this is a managed machine; there isn't a fix Hollowmark can apply from the daemon side.

  3. If Setup still shows "Keychain unavailable" after Credential Manager is confirmed working, right-click the system tray icon and choose "Try Again" — that's the daemon's retry control for this state (there's no equivalent action on the in-app Setup page). If it still doesn't clear, use the in-app bug report or Discord #bugs (see "Still stuck?" below) so we can look at it directly. The in-app uninstall's wipe option does not clear the Windows stored credential, so reinstalling that way won't fix this. Your account and data aren't affected.

Recovery — the short version (either platform)

  1. Quit the Hollowmark daemon and relaunch it.

  2. Open Hollowmark and check the Setup page. If it now shows Connected, you're done.

  3. Still showing Keychain unavailable? Click the menu bar icon (macOS) or system tray icon (Windows) and choose "Try Again."

  4. Still stuck? On macOS, reinstall the daemon (Settings → Danger Zone → Uninstall Daemon — tick "Also wipe my local config + cached data (irreversible)" before confirming, then reinstall). This resets the local credential store — you'll sign in again, but nothing else about your account or data is affected. On Windows, go straight to "Still stuck?" below and report it. Exception (macOS): if your symptom matches the one-time upgrade-migration case (see macOS causes above), reinstalling won't clear it — go straight to "Still stuck?" below and report it instead.

We intentionally don't ask you to run any terminal commands or edit any files to fix this — if the steps above don't resolve it, that's a signal something needs a closer look, not something to work around by hand.

Still stuck?

  • Use the in-app bug report (the Report a bug button in the app header). It attaches diagnostic logs automatically, which is the fastest way for us to see exactly what your daemon saw.

  • Or post in Discord #bugs with your OS/version and what you've already tried from this page.