Ticket Response

GPTClaudeDeepSeek··▲ 1,154 copies·updated 2026-07-14

ticket-response.prompt

---
title: Support — Ticket Response
owner: unassigned
status: populated
volatility: stable
last_reviewed: 2026-06-28
inject_into: [support-workflows]
tokens_approx: 0
---

# Support — Ticket Response

Task: draft a response to an inbound support ticket. Tone: calm and procedural first, empathetic second (`voice.md`). Lead with the fix.

## Steps

1. **Classify** the issue: down / degraded / consumable / training (`support/services.md`).
2. **Respond to class:**
   - **Down:** acknowledge impact, confirm we're acting, state the next step per the account's plan. Never invent a response time (RULE-02).
   - **Degraded:** offer the safe first-line check or a dial-in/training path before dispatch.
   - **Consumable:** route to operations replenishment; confirm the subscription cadence.
   - **Training:** offer the dial-in/training path; don't dispatch a tech for a skill gap.
3. **Set the next step** clearly: self-serve check, dispatch, or hand-off.

## Rules

- Never walk a customer through an unsafe (electrical/plumbing) repair — dispatch a tech.
- No invented SLAs, prices, or specs (RULE-01/RULE-02).
- If the fix isn't grounded in context, flag `[REVIEW: ...]` and route, don't guess.

## Output

A send-ready customer reply that leads with the fix or the next action.

when to use it

Community prompt sourced from the open-source GitHub repo matthewjbis/context-kit (no explicit license). A "Ticket Response" style prompt — adapt the placeholders and specifics to your task. Imported as-is and not independently retested here, so check the output before relying on it.