[codex] fix SupportOps Supabase recovery

[sonnymay]

Summary

• Add safe Supabase request handling so database failures return clear 503 responses instead of raw 500s.
• Add /health/dependencies for Render/Supabase readiness checks.
• Convert blank optional relationship IDs from the frontend to null before sending UUID fields to Supabase.
• Add backend/supabase/schema.sql so SupportOps tables and demo rows can be recreated from source.
• Document the dependency healthcheck and schema file.

Root Cause

Production /health worked, but /dashboard, /customers, and /tickets returned 500 because the Supabase supportops project was inactive and the public schema had no usable app data. I restored the Supabase project, verified it is healthy, and seeded/recovered the app tables.

Validation

• Backend live endpoints recovered before code deploy: /dashboard, /customers, /tickets, /devices, /rmas returned 200.
• Local backend tests: 22 passed.
• ruff check, ruff format --check, and mypy passed.
• Frontend build passed.

Deploy Notes

After merge/deploy, verify /health/dependencies and the main API routes on Render.

[vercel]

The latest updates on your projects. Learn more about Vercel for GitHub.

Project	Deployment	Actions	Updated (UTC)
supportops	Ready	Preview, Comment	Jun 16, 2026 4:43am