tuist install blocks for several minutes on “Uploading run metadata…” when server errors

What happened?

In CI, tuist install finishes package resolution but then hangs on Uploading run metadata… for several minutes before finally failing with a 500.

Removing fullHandle from the config eliminates the hang entirely (no upload is attempted).

The server is self-hosted via TUIST_URL.

Expected behavior tuist install should not block command completion on best-effort analytics upload. It should upload in the background (as it does locally) and/or be bounded by a short overall timeout and a sane retry budget, so a slow/erroring server can never stall a finished command.

Actual behavior The command blocks on a synchronous analytics upload for minutes(~8 minutes) due to compounding retries, then fails.

ref: [TrackableCommand.swift] (line : 183)

How do we reproduce it?

1. Configure Config(fullHandle: “/”) with a self-hosted server (TUIST_URL) that returns 5xx or is slow/unreachable.
2. Run tuist install in CI (Environment.current.isCI == true).
3. After dependencies resolve, the command prints Uploading run metadata… and blocks for minutes(~8 minutes), then logs Failed to upload run metadata: unknownError(500).

Error log

Resolving and fetching plugins.
Resolving and fetching dependencies.
Uploading run metadata...
Failed to upload run metadata: unknownError(500)

macOS version

26.5.1

Tuist version

4.201.0-canary.12

Xcode version

26.0

No GitHub comments yet.