[Bug] Anthropic API Error: Overly broad safeguards flagging legitimate code review requests
Status Open
Reported on v2.1.220
Maintainer reply None cached
Activity 0 comments · opened Aug 4, 2026
Bug Description
You flagged legitimate coding efforts. I asked Fable 5 for an "adverserial review", because that's what Opus 5 called it, of the work that Opus 5 did for me in a branch. It was midstream and I got the safeguard message "Fable 5's safeguards flagged this message. Our intentionally broad safeguards allow us to deliver more capabilities faster, but can sometimes
flag legitimate coding, cybersecurity, and biology tasks. Switched to Opus 5. Send feedback with /feedback or learn more"
Environment Info
- Platform: darwin
- Terminal: iTerm.app
- Version: 2.1.220
- Feedback ID: 1c5a0dcc-35d8-4eab-b182-5f47c3319607
Errors
[{"error":"Error: 500 {\"type\":\"error\",\"error\":{\"type\":\"api_error\",\"message\":\"Internal server error\"},\"request_id\":\"req_011CdWvxrkgZKF1H5LkCfbfw\"}\n at generate (/$bunfs/root/src/entrypoints/cli.js:40:50094)\n at makeRequest (/$bunfs/root/src/entrypoints/cli.js:80:7690)\n at processTicksAndRejections (native:7:39)","timestamp":"2026-07-29T19:51:35.376Z"}]