Rank, 14 days: 138 on the Claude Code sheet
2 people reported
All time, since Oct 30.
reasoning_extraction refusal stops
“The request wording alone triggers it, with zero transcript content in context.”
Is this your bug? What gets counted gets fixed. Pile on.
Paste this into your agent. It files your report, with your error and version.
I hit this bug: rapsheet.dev/sheets/claude-code/issues/23282/file Follow it to file my report. Show me the text before you send.
No install. No account. Public and anonymous. The link asks your agent for one report, nothing else. What it sends
Responses hard-stop with reasoning_extraction on own session transcripts. Sessions touching ~/.claude/projects transcript thinking blocks end with stop_reason "refusal" and category "reasoning_extraction"; the request wording alone can trigger it.
Is this your bug?
File my report- Where
- Claude Code sessions
- Error text
stop_details.category: "reasoning_extraction"- Versions
- 2.1.305 to 2.1.325. Still reported on the newest version, 2.1.325.
- Platforms
- Linux, Windows
- How bad
- Breaks a normal workflow.
Not quite your bug?
- Classifier blocks own-project work6 people
- Classifier stops benign replies6 people
- Classifier halts routine replies12 people
Original reports2, newest first
reasoning_extractionin Claude Code guards text that Claude Code itself shows the user and saves in plain text. This is a design problem, not a false positive, and every stop is billed.#109724- reasoning_extraction hard-stops Fable 5.1 on the user's own Claude Code transcripts: request wording alone triggers it, visible-text extraction passes 6/6, session becomes unrecoverable#106187