[Feature Request] Refine Fable 5 safeguards to reduce false positives on legitimate cosmetics/skincare app development

Bug Description
Fable 5's safeguards produce false positives on legitimate skincare / cosmetics app development. I'm a paying Max subscriber (~$200/month for many months) building a non-MDR iOS skincare app — appearance-only, 100% on-device, zero medical or diagnostic claims.

Routine, safe development tasks are repeatedly flagged as unsafe "biology" work and auto-switched to Opus 4.8: analyzing my own skin-imaging Swift code, reviewing color/texture/erythema image-processing methods, discussing dermatology-adjacent appearance metrics for a cosmetic app. None of it is dual-use or harmful — it is consumer cosmetic-appearance software, developed legally.

The flag message itself admits the safeguards are "intentionally broad" and flag safe coding / biology work. Understood — but the practical effect is that a loyal paying developer is blocked from using your newest, best model on the exact work it excels at (orchestration, reasoning), and pushed back to an older model. That is a real productivity and trust cost.

Request: refine the safeguards so legitimate cosmetics / skincare / derm-adjacent app development is not caught, and/or give paying users a faster false-positive review/allowlist path. Over-braking new-model access for legit paying customers erodes trust fast.

Environment Info

• Platform: darwin
• Terminal: Apple_Terminal
• Version: 2.1.197
• Feedback ID: debafdc0-8030-4cc6-9ab2-1dd1df97247d

Errors

[]