[Bug]: Android T3 Account stays Checking and all T3 Connect environments fail until app relaunch #24306

Before submitting

• I searched existing issues and did not find a duplicate. Related reports are listed below.
• I included enough detail to reproduce or investigate the problem.

Area

apps/mobile

Steps to reproduce

Observed workflow, not a deterministic reproduction:

1. Have previously configured T3 Connect remote environments and a separate direct/Tailscale environment in the native Android app.
2. Open the app. In the reported episode, none of the three T3 Connect environments could connect.
3. Open Settings → Environments: all three cloud environments have red indicators and connection errors, although their switches are enabled. The direct/Tailscale environment is green and says “Connected”.
4. Open Settings: the “T3 Account” row is greyed out with status “Checking”.
5. The reporter closed the Android application, waited for a while, and reopened it; connectivity then returned. The exact wait and whether this was an OS force-stop or ordinary app dismissal were not captured.

The initial trigger is unknown. No network disruption, account change, or background-duration condition was established.

Expected behavior

The Android client should finish account initialization and connect to saved T3 Connect environments, or show an actionable initialization error/retry state. A recoverable condition should not require closing the whole application and waiting before reopening it.

Actual behavior

• Screenshot of Settings: “T3 Account” is greyed out and displays “Checking”.
• Screenshot of Environments: one direct/Tailscale environment says “Connected”; all three T3 Connect environments are red with their switches enabled.
• Each visible error is truncated to “Connection failed. Reason: Your clo…”. The full error was not expanded/captured. This report does not assume it was the complete “Your cloud sign-in changed” error or diagnose authentication from the truncated string.
• The reporter could not connect to the T3 Connect environments until closing the app, waiting, and reopening it.

The Settings and Environments screenshots were captured separately. They are not a synchronized trace, and do not establish the duration of the “Checking” state. Direct connectivity at capture time does not prove that every cloud dependency was reachable.

Impact

Major degradation or frequent failure

Remote work through all configured T3 Connect environments was unavailable during the observed episode. The direct environment remained connected in the supplied screenshot. Recurrence frequency and exact outage duration are unknown.

Version or commit

Affected main remote server: v0.1.0, established during the preceding investigation. Android app version: 2.0.0, reported by the user. Build number: not yet captured. The exact versions of the other cloud environments were not rechecked for this report.

Environment

Native Android app. Three T3 Connect environments plus one direct HTTPS/Tailscale environment. Exact phone model, Android OS version, app build, network-transition history, and cloud-account initialization logs are not yet captured.

Logs or stack traces

No Android logs or cloud-auth/network stack trace was captured. No proof of a Clerk load failure, expired token, server outage, VPN fault, or schema mismatch is available.

Screenshots, recordings, or supporting files

GitHub-hosted copies of the reporter's actual screenshots. Private labels/hostnames are redacted. The PNG screenshots were losslessly converted to self-contained SVGs and independently rendered back to PNG for a pixel-for-pixel equality check; no UI content was generated.

Workaround

Close the Android application, wait for a while, then reopen it. This restored connectivity in the reported episode. The exact wait and dismissal/force-stop method are unknown; this is an observed recovery, not a guaranteed workaround. No sign-out, credential clearing, environment deletion, re-pairing, or server restart was reported as necessary.

Related reports and evidence limits

• [Bug]: Account controls and T3 Connect stay unavailable after Clerk UI load failure until Reload #15725: desktop/macOS account controls and T3 Connect unavailable after a confirmed Clerk UI load failure, recovered by Reload. Similar recovery pattern, but this Android case has no Clerk error log and should not inherit that diagnosis.
• [Bug]: T3 Connect reports "Your cloud sign-in changed" when the host is just overloaded #11004: a cloud sign-in error can be misleading; the full error here is not captured.
• [Bug]: Android t3 connect 2.0.0 cant connect to desktop v0.1.0 #21206: Android 2.0.0 / server v0.1.0 connection failure, but that report describes the connection option being unavailable, rather than existing enabled environments failing while the account is Checking.

This is separate from the report about an individual conversation not rendering after a provider switch. No shared root cause is established.

Prepared by GPT-6.1-Sol through Hermes Agent on behalf of the reporter. No credentials, private conversation contents, or private hostnames are included.