Skip to main content

Production Journey Tests — End-to-End on Live Sites

non-critical   Property: ChurchWiseAI   Category: UX / Flow Tier: all Persona: anonymous Touchpoint: production URLs (churchwiseai.com, pewsearch.com, illustratetheword.com)

Preconditions

  • All 3 production sites deployed and live
  • Demo church data loaded in Supabase
  • Anthropic API configured

Steps

#ActionExpected Result
1Load CWA care hub for demo churchPage loads. Care agent cards visible. At least 1 agent card with clickable link.
2Click care agent cardNavigates to chat page for that agent.
3Load CWA chat pageChat interface loads. Input field and send button visible.
4Send test message to chatbot APIPOST /api/chatbot/stream returns 200. Response contains AI-generated text (>10 chars). Uses HEAR protocol.
5Load PewSearch homepageDirectory search page loads. Church listing/search visible.
6Load ITW illustration browseIllustration grid loads. At least 1 illustration card visible.
7Check all sites for console errorsNo JavaScript errors in browser console. Network requests successful.
8Test mobile viewport on CWAHomepage responsive. Chat widget adapts to mobile. No layout breakage.

Known Failure Modes

  • Chat API returns 500 — LLM or Supabase issue
  • Care hub shows no agent cards — Supabase query failing
  • Chat interface doesn't load — missing JS dependencies
  • Mobile layout broken — responsive design issue

References

Notes

Tests real user journeys on LIVE production URLs — not just page loads, but actual interactions: clicking cards, sending messages, getting AI responses. Uses demo church in production. Tests that Anthropic API, Supabase queries, and frontend interactivity all work end-to-end. Run after deployment to verify the full stack is operational.