[Bug]: T3 Connect managed endpoint provisioning fails on every startup retry (Windows 11 Pro, 0.0.42): managed_endpoint_provisioning_failed / HTTP 504 #14155

Before submitting

• I searched existing issues. This looks like the same relay-side failure as [Bug]: T3 Connect managed endpoint provisioning fails repeatedly on Windows 11 #14070 (open) and [Bug]: T3 Connect managed endpoint provisioning fails for Windows environment #13681 (closed after a later retry succeeded). I'm filing it separately with this environment's IDs so the relay logs can be checked.
• I included enough detail to reproduce or investigate the problem.

Area

apps/server

Steps to reproduce

1. On Windows 11 Pro, set up T3 Connect from the desktop app. The relay client (cloudflared 2026.5.2, managed install) installed fine, and the CLI authorization is stored and valid.
2. Fully restart T3 Code so the server runs the startup link reconcile (environment.cloud.reconcileDesiredLink).
3. Open the T3 Code mobile app (Android) → Environments.

Expected behavior

The relay provisions the managed endpoint and returns a tunnel token. cloudflared starts, and the phone can open the environment.

Actual behavior

Every POST https://relay.t3.codes/v1/client/environment-links from the startup reconcile fails, for the full 10-minute retry window. Responses alternate between managed_endpoint_provisioning_failed and a bare HTTP 504 from Cloudflare. No tunnel token comes back, so cloudflared never starts.

The environment still shows up on the phone under T3 Connect, but opening it fails with:

Failed to connect. Reconnecting... Reason: Relay could not reach the environment endpoint (endpoint_request_failed). Trace ID: c558438234c328c6e3cb1279159b47f2

Checks done on this machine:

• relay.t3.codes resolves and is reachable (HTTPS 200).
• POST /v1/client/environment-link-challenges with the stored CLI token returns 200, so auth and network are fine.
• The link proof is generated successfully (environment.cloud.makeLinkProof succeeds on every attempt).
• cloudflared.exe version runs fine (2026.5.2).
• t3 connect status before the restart: Exposure enabled, Authorization stored credential, Environment link pending server startup, Relay client available via managed install.

Environment ID: 1408daaf-57fe-4414-89fd-b811a2d5befa

Could someone check relay.managed_endpoint_provider.provision for these traces?

Mobile-side trace: c558438234c328c6e3cb1279159b47f2

Impact

Blocks work completely

Version or commit

T3 Code (Alpha) desktop 0.0.42 (Windows installer)

Environment

Windows 11 Pro 10.0.26200 (x64), desktop app 0.0.42, managed cloudflared 2026.5.2, Android mobile app. Region: South Africa (Cloudflare JNB edge).

Logs or stack traces

EnvironmentHttpInternalServerError: T3 Connect: Relay environment link is unavailable: managed_endpoint_provisioning_failed. Retry when the relay is available. If this continues, include the trace ID when reporting it. Trace ID: a3f31bc2f7f71f2a72a82af1013d9ab6.
    at environment.cloud.reconcileDesiredLinkWith (bin.mjs:205751:16)
    at environment.cloud.reconcileDesiredLink (bin.mjs:206157:11)

EnvironmentHttpInternalServerError: T3 Connect relay returned HTTP 504 without a recognized error response. Check relay access and any proxy or firewall restrictions, then restart T3 Code. Cloudflare Ray ID: a42551115fd07885-JNB.
    at environment.cloud.reconcileDesiredLinkWith (bin.mjs:205751:16)
    at environment.cloud.reconcileDesiredLink (bin.mjs:206157:11)

Workaround

None found yet. Restarting T3 Code re-runs the reconcile, but every attempt in the retry window failed.