Its theme's rank, 14 days: 2 on the Claude Code sheet
48 people reported
All time, since Sep 24, including 3 thumbs-up on GitHub.
Safeguards refuse work requests
“The safeguards, they have blocked my general work and I'm not able to continue.”
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/1699/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
Safeguards flag and refuse general work requests. Safety checks catch random requests and keep flagging whatever the user says, blocking them from continuing their work.
Is this your bug?
File my report- Where
- safety checks
- Versions
- 2.1.236 to 2.1.319. Still reported on the newest version, 2.1.319.
- Platforms
- Linux, macOS, Windows
- How bad
- Some reports say it blocks them completely.
Part of: Safeguards block session requests34 bugs
- Cyber safeguard blocks sessions377 people
- Reasoning_extraction safeguard block134 people
- Cyber safeguard pauses sessions52 people
- Safeguards block requests47 people
- Classifier halts routine replies12 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 reports46, newest first
- User incorrectly flagged for cyber misuse with no apparent cause#108234
- Chat incorrectly flagged benign container work as cybersecurity misuse#108125
- Safeguard Refuses Benign PS5 Homebrew Temperature Monitor Request#108119
- Benign coding requests incorrectly flagged by cyber safety filter#108087
- Safety classifier blocking requests in session and predecessor with no apparent cause#107964
- Fable safeguards block benign prompts with false positives#107603
- Opus 5.5 flags benign prompts (even "hi") as policy violations#107386
- Excessive safeguard errors blocking legitimate work in own project#107078
- Legitimate SDK bridge work for image editors incorrectly flagged as cybersecurity#106969
- Legitimate infrastructure task request incorrectly flagged#106963
Show all 46Show only the newest 10
- Non-malicious request incorrectly blocked as hacking#106833
- Claude Code refuses legitimate marketing and development prompts as security violations#106254
- Safety filter triggers on benign requests with no apparent cause#106249
- Legitimate requests incorrectly flagged as violations#106031
- User Message Blocked by Content Filter Before Reaching Model#105798
- Safeguards Incorrectly Flagging General Work Requests#105790
- Safeguards flagging and refusing benign requests#105565
- Requests repeatedly flagged#105485
- False Positive Content Flagging on Benign Input#105294
- Claude Code flags conversation as unsafe and blocks further work#105215
- Safeguards incorrectly flag benign portfolio website request#105078
- Request Unexpectedly Flagged#104499
- Bug Report: Claude Code refuses requests and gives unhelpful responses#104440
- False positive flagging during legitimate operations#104424
- Non-cybersecurity request incorrectly flagged as cybersecurity#104242
- Excessive Safety Refusals on Basic Requests#104055
- y[Bug] Unexpected refusal on benign requests across multiple prompt variants#103984
- Cyber guard frequently blocks benign requests in session#103163
- Inappropriate Safety Filter Flagging Non-Safety Content#103079
- Overly broad blocking of normal research and codebase questionsg#103042
- Safety checks incorrectly refuse benign requests#102851
- Excessive safeguards blocking legitimate work tasks#102450
- Claude Code refuses GUI-building task citing safeguards despite benign intent#100467
- Claude Code blocked request due to safety system violation#100137
- Auto-mode safety classifier incorrectly blocks legitimate emulsion engineering tasks and persists across session#99433
- Security filter incorrectly blocks legitimate FiveM scripts#99106
- Anthropic API Error: False-positive cyber refusal on basic arithmetic requests in research workspaces#99037
- Excessive safety checks blocking code generation workflow despite being in non-productioin mode.#98666
- Anthropic API Error: Request Blocked#98626
- Legitimate statusline modification incorrectly flagged as safeguard violation#98558
- Safety guardrails blocking legitimate game development use cases#98258
- Safety classifier blocking legitimate operations during analysis and UI rendering#98203
- Anthropic API Error: Safeguard Policy Blocking Legitimate App Development Use Case#97809
- Anthropic API: Safety Classifier False Positive on Routine Engineering Design Task#97743
- Anthropic API Error: Request blocked by safety classifier#97570
- [Bug Report: Unable to process request, because of safeguard. On Sonnet!!#96667