Daily Error Log Summary

You are generating the daily error log summary. Review all application error output from the past 24 hours and produce a concise briefing for the engineering team. ## Log Sources Check the following locations for error data (use whichever are present in the environment): - Application logs: `/var/log/app/error.log` or `./logs/error.log` - Next.js / Node.js stderr output captured in the deployment platform logs - Vercel function logs (via `vercel logs --since 24h` if CLI is available) - Any structured JSON log files under `./logs/` with a timestamp within the past 24 hours If log access requires a command, run it. If logs are unavailable, note that explicitly and stop. ## Steps 1. **Collect errors from the past 24 hours.** Filter for log entries at level `ERROR`, `FATAL`, `CRITICAL`, or HTTP status >= 500. Parse timestamps to confirm the 24-hour window. 2. **Deduplicate and count occurrences.** Group identical or near-identical error messages together. Track: error message (normalized), first seen, last seen, count. 3. **Categorize each error group** into one of these types: - **Application crash / unhandled exception** — process exited or threw uncaught error - **Database / query error** — connection failures, timeout, constraint violations - **External API / third-party failure** — upstream service errors, 5xx from external calls - **Authentication / authorization error** — 401/403 errors beyond normal user mistakes - **Not Found / routing error** — unexpected 404s that may indicate broken links or bad deploys - **Rate limit / throttle** — 429 responses from external services - **Validation / input error** — 400 errors or schema validation failures - **Unknown** — anything that doesn't fit above 4. **Identify recurring patterns.** Flag any error group that: - Appeared more than 10 times in 24 hours - Is new (not seen in the prior 7 days, if a baseline exists) - Shows an increasing frequency trend within the 24-hour window 5. **Check for critical signals.** Immediately escalate if any of the following are found: - An unhandled exception or process crash - Database connection failures sustained for more than 5 minutes - Any error affecting the payment or authentication flow - Error rate spike: more than 100 errors in any single 1-hour window ## Output Format ### Date: [YYYY-MM-DD] | Window: Last 24 hours ### Overall Health: HEALTHY / DEGRADED / CRITICAL ### Error Summary Table: Category | Unique Error Groups | Total Occurrences | Trend (↑ / → / ↓) ### Top 10 Errors by Occurrence Table: # | Error Message (truncated) | Type | Count | First Seen | Last Seen | New? ### Recurring / Spike Alerts If none: "No spikes or new recurring errors detected." If any: numbered list with error, count, window, and recommended action. ### Critical Escalations If none: "No critical issues requiring immediate action." If any: bold header "ACTION REQUIRED" followed by description and on-call notification recommendation. ### Notes Any context that would help the team understand the errors (e.g., correlates with a deploy at 14:32, elevated traffic from a marketing campaign, etc.).