Proto Production Debugging

Diagnose Proto production failures across Vercel, Supabase, catalogue integrations, stock, registration, uploads, and order flows.

Gee9999 Updated

File contents

Use this skill for Proto production incidents and reliability work. Read before changing code:

  1. Reproduce the failure against the relevant preview or production route when safe.
  2. Inspect Vercel runtime errors and deployment metadata.
  3. Trace the route into its integration boundary: Supabase, catalogue, stock, Instore, Brevo, or authentication.
  4. Separate user-input errors, transient upstream failures, configuration failures, and code defects.
  5. Write a failing regression test before implementing a fix.
  6. Add bounded timeouts, useful error categories, and safe user-facing fallbacks where appropriate.
  7. Verify the complete flow in a browser and inspect the post-fix runtime signal.

Known Proto signals to check first include shared url.parse() deprecation warnings, incomplete extended-range catalogue lookups, stock timeouts, missing Instore configuration, invalid catalogue credentials, unsupported .jfif uploads, weak-password registration errors, and idempotent archive/unarchive actions.

Never expose secrets, service keys, auth tokens, or private customer data in logs, issues, screenshots, or commits. Do not suppress a warning until its source is identified.

Gee9999/codex-skills/tree/main/proto-production-debugging commit 54057ae291

Frequently asked questions

npx skillmds@latest add gee9999/proto-production-debugging