[BUG] Safety classifier too sensitive - switches to Opus 4.8 for harmless commands like /feedback

Resolved 💬 2 comments Opened Jul 12, 2026 by donlagon1 Closed Jul 15, 2026

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?

Fable 5's safety classifier is too aggressive. Even typing harmless commands like /feedback triggers an automatic switch to Opus 4.8, wasting tokens. This happens even with autoSwitchModels: false in settings.json.

What Should Happen?

Fable 5 should handle simple commands without flagging them as unsafe. The safety threshold should be more reasonable.

Error Messages/Logs

Steps to Reproduce

  1. Set autoSwitchModels: false in ~/.claude/settings.json
  2. Open Claude Code and select Fable 5 as the model
  3. Type /feedback
  4. Claude Code switches to Opus 4.8 automatically

Claude Model

None

Is this a regression?

Yes, this worked in a previous version

Last Working Version

_No response_

Claude Code Version

Claude 1.20186.1 (df1d8a) 2026-07-10T21:55:12.000Z

Platform

Anthropic API

Operating System

macOS

Terminal/Shell

Terminal.app (macOS)

Additional Information

_No response_

View original on GitHub ↗

This issue has 2 comments on GitHub. Read the full discussion on GitHub ↗