Monitoring and Tracking
Sentry catches errors, Vercel keeps the logs, Linear holds the issues and in-app reports.
Sentry
What it does. Receives every uncaught error from pages, server actions and API routes in production, plus an explicit report when an application fails to save. Check it first when someone says "it errored".
Configuration. NEXT_PUBLIC_SENTRY_DSN. Enabled only when the app runs in production. Source maps are not uploaded yet, so stack traces show built file names; adding withSentryConfig to next.config.ts would fix that.
Vercel Logs
Every request, and the console output of every scheduled job. Filter by route to see what a job did this morning. Build logs are under each deployment.
Clerk Logs
Sign-in attempts, invitation states, and the webhook delivery log with the exact response the platform gave.
Resend
Delivery, bounces and opens per message. Suppressed addresses.
Supabase
Database health, slow queries, backup status, and cron.job_run_details for the five-minute job.
Linear
The issue tracker for the build and for support. The Report a problem dialog in the app files an issue here with the page, screen size, browser, recent console errors and a screenshot, and emails a copy to the Impact Hub team. The workspace transfers to Impact Hub Houston at handover, so the history stays with you. Configured by LINEAR_API_KEY, LINEAR_TEAM_ID and LINEAR_SUPPORT_LABEL_ID; without them, reports go by email only.
What Is Not There Yet
No uptime monitor and no health route. A free external check against https://www.impacthubos.com/sign-in every few minutes is a worthwhile addition.