Its theme's rank, 14 days: 2 on the Claude Code sheet
12 people reported
All time, since Sep 23.
Classifier halts routine replies
“What happens: Many responses are cut with "Your response above was stopped by a safety classifier — this is not a tool or API error".”
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/11379/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
Safety classifier halts replies on routine, non-sensitive work. Replies are cut with "Your response above was stopped by a safety classifier — this is not a tool or API error", withholding the rest of the text.
Is this your bug?
File my report- Where
- Response safety classifier
- Error text
Your response above was stopped by a safety classifier — this is not a tool or API error.- Versions
- 2.1.278 to 2.1.319. Still reported on the newest version, 2.1.319.
- Platforms
- Linux, macOS, Windows
- How bad
- Breaks a normal workflow.
Part of: Safeguards block session requests34 bugs
- Cyber safeguard blocks sessions377 people
- Reasoning_extraction safeguard block134 people
- Cyber safeguard pauses sessions52 people
- Safeguards refuse work requests48 people
- Safeguards block requests47 people
- Requests blocked with a [cyber] flag8 people
- Refusal on own-code hardening11 people
- SSH to own devices refused7 people
- Pentest work blocked6 people
- CTF requests flagged as malicious6 people
Show all 33Show only the largest 10
- Classifier blocks own-project work6 people
- Classifier halts reverse engineering5 people
- Safeguards flag verified cyber work6 people
- Firmware requests flagged as cyber6 people
- Classifier stops benign replies5 people
- reasoning_extraction API error5 people
- Safeguard blocks all messages4 people
- [bio] classifier blocks responses4 people
- Reasoning extraction safeguard block4 people
- Security investigation requests blocked3 people
- Classifier stops DeFi bot work3 people
- Safeguards block CVP members3 people
- "[cyber]" flag stops sessions1 person
- Safety classifier blocks agent3 people
- Fictional hacking content flagged3 people
- Cyber verification prompt recurs2 people
- Cyber safeguards block own devices2 people
- Safety classifier stops auth work2 people
- Phishing research project flagged2 people
- Cyber stop on report tasks2 people
- Cyber safety block on audits2 people
- Biological safeguard triggers in sessions with no biology content2 people
- Unexpected safeguards notice2 people
Original reports12, newest first
- Safety Classifier False Positive Cuts Defensive auditd Sensor Code Responses#108061
- Safety classifier false positive blocks routine token-cost analysis in Claude Code desktop#106876
- Safety filter stops read-only inventory and routine replies as false positives#105561
- Safety classifier stops harmless responses (false positives), e.g. a plain "you're welcome"#104983
- Safety classifier repeatedly stops replies on ordinary project work (hundreds of times over 3-4 days)#101974
- Safety classifier withholds responses during open-source automotive (openpilot fork) work#101967
- Safety classifier stops defensive sandboxing work (seccomp/Landlock) as false positive#101932
- Safety classifier withholds responses during legitimate healthcare documentation work#100212
- Response-level safety classifier false-positive halts a completely benign reply (harness-generated session-confirmation turn)#98556
- Safety filter blocks legitimate internal documentation/demo content#98510
Show the newest 11 of 12Show only the newest 10
- Safety Classifier Issue#96236