Fix Authentication & Logic for AI Scam Detector App

I have a web app (AI scam detector) built with a no-code/AI builder and I need help fixing the authentication and scan flow.

Current issues:
• Signup shows “failed to create record” even when successful
• First scan result does not display after signup
• Scan counter is out of sync (top header shows incorrect value)
• State is not persisting correctly after authentication

What I need fixed:
1. Clean signup/login flow with no false errors
2. After signup, immediately run and display the first scan result
3. Implement a single source of truth for scan_count
4. Ensure both header and result modal show the same remaining scans
5. Ensure scan input persists through signup flow

This is NOT a full rebuild — just fixing logic and state handling.

Tech:
• AI builder with integrated backend (Horizons-style)
• Basic auth + database already set up

Goal:
Fully working flow:
User enters scan → signs up → sees result → scan count updates correctly → paywall after 3 scans

Please confirm:
• You understand async auth flows
• You can fix state management and UI sync issues
• Estimated time to complete

I expect this to be a quick fix for someone experienced.