[Bug]: Repos without a remote record a failed resolvePrimaryRemoteName span on every status refresh #32904

Before submitting

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

Area

apps/server

Steps to reproduce

1. Make a git repo with one commit and no remote, on a branch without an upstream:

   git init demo && cd demo
   git commit --allow-empty -m init
   git checkout -b feature

2. Add demo as a project in T3 Code and open a thread in it.
3. Wait a few status refreshes (the automatic git fetch interval, 30 s by default).
4. Open Settings → Diagnostics.

Expected behavior

No failures. A repo without a remote is a normal setup, and status works fine for it.

Actual behavior

resolvePrimaryRemoteName shows up as a failure, and the count goes up on every refresh (two failures per status call on a feature branch).

Status itself is correct. The callers catch the error and fall back to the local branch. Only the span is wrong: the "no remote" error is raised inside the traced Effect.fn, so the span ends as a failure even though every status caller expects it.

• The span and the error: GitVcsDriverCore.ts#L1741, #L1750-L1757
• Callers that recover with orElseSucceed: GitVcsDriverCore.ts#L1907, GitManager.ts#L1429, GitManager.ts#L2077
• Refresh interval default: settings.ts#L1001

Over time these expected failures crowd out real ones in Diagnostics.

#29070 fixes the same pattern for a different probe.

Impact

Minor bug or occasional failure

Version or commit

main @ e11c2fd (also seen on 0.1.7)

Environment

Linux x64, server

Logs or stack traces

GitCommandError: No git remote is configured for this repository.
  operation: GitVcsDriver.resolvePrimaryRemoteName, args: ["remote"]

Workaround

None needed for behavior. Adding any remote stops the failures.

Written by Claude Opus 5.5 (effort: high) in Claude Code.