[BUG] Remote triggers with git_repository source lose GitHub authentication and require manual re-authorization

Preflight Checklist

• [x] I have searched existing issues and this hasn't been reported yet
• [x] This is a single bug report (please file separate reports for different bugs)
• [x] I am using the latest version of Claude Code

What's Wrong?

Description

Remote triggers configured with a git_repository source periodically lose their GitHub authentication, causing runs to fail. The failure requires manually re-authorizing GitHub in claude.ai settings and updating the trigger to pick up fresh credentials.

What Should Happen?

Remote triggers with git_repository sources should automatically refresh GitHub tokens between scheduled runs without manual intervention.

Error Messages/Logs

Two failure modes observed:

  1. Setup failure:
  ▎ An unexpected error occurred while setting up the git_repository source. Try starting a new session. If the problem persists, the error details have been logged — contact support with your session ID.
  2. Interactive auth fallback (impossible in headless environment):
  ▎ To fetch the failure report from the dbt-monitoring branch, I need access to GitHub. Please open this URL in your browser to authorize...

Steps to Reproduce

Environment

• Trigger type: Scheduled (cron)
• Source: git_repository with private GitHub repo
• Schedule: Daily (0 14 *)

Claude Model

Opus

Is this a regression?

I don't know

Last Working Version

_No response_

Claude Code Version

2.1.111

Platform

Anthropic API

Operating System

macOS

Terminal/Shell

VS Code integrated terminal

Additional Information

Current workaround

Re-authorize GitHub in claude.ai settings, then update the trigger via RemoteTrigger API to pick up fresh credentials. This has to be repeated each time the token expires.