[Bug] Anthropic API Error: Fable 5 safeguards falsely flagging safe messages

**Bug Description**
fable is currently unusable because with every second message it gives me this error. Even the safest prompts with no exploitation or whatever! API Error: Fable 5's safeguards flagged this message (https://www.anthropic.com/legal/aup). This sometimes happens with safe, normal conversations. Claude Code can't respond to this message with Fable 5.

Double press esc to edit your last message, or try a different model with /model.

Send feedback with /feedback or learn more: https://support.claude.com/en/articles/15363606

Details: `[reasoning_extraction]`

Request ID: req_011CeDpLqt3THEwBYjQqWnGh. I get this in all prompts be it a web search, changing things via api in gitlab etc.!!!

**Environment Info**
- Platform: darwin
- Terminal: xterm-256color
- Version: 2.1.237
- Feedback ID: 34d5bb22-82d3-40c8-b0ab-62e0c1e5a784

**Errors**
```json
[]
```

Reproduce the reported safeguard error in Claude Code 2.1.237 using Fable 5 and a safe prompt, then use the supplied request ID and feedback ID to investigate the failure. Done means determining whether Claude Code contributes to the false flag and documenting or correcting a reproducible issue.

索引モデルが issue の本文から書いたものです。