09: Build-Along - Support Ticket Triage + Queue Digest

Goal: automate daily support triage so humans do less routing and more resolution.

You will implement one practical workflow:

1. read incoming ticket objects,
2. route each ticket into explicit queues,
3. emit a digest email for support leads.

Why This Matters

Most teams still triage support tickets manually in Slack/Jira threads.

This workflow creates one reviewable source of truth for routing policy.

Final Workflow File

• examples/realworld/support-ticket-triage-digest.gr

Business Rules

Current routing logic:

• critical -> incident-oncall
• high -> priority-response
• everything else -> standard-backlog

Each line includes:

• ticket id
• chosen queue
• priority
• customer tier
• SLA minutes
• topic

Run It

CLI path:

GRAPHEME_ALLOWED_SMTP_DOMAINS=example.com \
  grapheme run examples/realworld/support-ticket-triage-digest.gr --json

Workspace path:

GRAPHEME_ALLOWED_SMTP_DOMAINS=example.com \
  cargo run -- run examples/realworld/support-ticket-triage-digest.gr --json

Expected Output Signals

Look for:

• "outcome": "succeeded"
• SMTP accepted state
• digest body containing queue assignments for all tickets

Required Failure Drills

1. Remove SMTP allow-list env var.

• Observe failure and document remediation.

2. Add a new critical ticket row.

• Verify it routes to incident-oncall.

3. Downgrade a high ticket to normal.

• Verify route changes to standard-backlog.

Production Adaptation

1. Replace inline ticket list with your ticket source integration.
2. Add business-hour and region routing rules.
3. Add VIP customer overrides.
4. Add escalation summary metrics in email footer.

This is the practical pattern: explicit routing policy in code, predictable digest output for operators.