[BUG] Scheduled routine runs ignore configured model and fall back to Sonnet 4.6 — manual runs use correct model (Opus 4.8)

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?

Describe the bug

A Claude Code routine configured to use Opus 4.8 (claude-opus-4-8) consistently executes on Sonnet 4.6 (claude-sonnet-4-6) when triggered by the schedule. Running the same routine manually ("Run now") correctly uses Opus 4.8.

The model setting has not changed between manual and scheduled runs. The routine has not been re-saved since the model was set.

What Should Happen?

Expected behavior

Scheduled runs should use the model configured in the routine settings, same as manual runs.

Actual behavior

Scheduled runs silently downgrade to Sonnet 4.6 regardless of the configured model. The system context inside the session confirms the active model (claude-sonnet-4-6).

Error Messages/Logs

Steps to Reproduce

To reproduce

Create a routine at claude.ai/code/routines
Set the model to Opus 4.8
Save and configure a daily schedule trigger
Run manually → session executes on Opus 4.8 ✓
Wait for the scheduled run → session executes on Sonnet 4.6 ✗

Claude Model

Opus

Is this a regression?

No, this never worked

Last Working Version

_No response_

Claude Code Version

2.1.156 (Claude Code)

Platform

Anthropic API

Operating System

macOS

Terminal/Shell

Terminal.app (macOS)

Additional Information

Environment

Platform: Claude Code on the web (claude.ai/code)
Configured model: Opus 4.8 (claude-opus-4-8)
Actual model on scheduled run: Sonnet 4.6 (claude-sonnet-4-6)
Trigger: daily schedule (8:00 AM America/New_York)
Manual runs: unaffected (use Opus 4.8 correctly)