Complete AI Training

Prompt

Update A Canned Response Library

Use this when you need canned responses updated and organized so agents answer common questions consistently.

How to use it

  1. Copy the prompt and paste it into ChatGPT, Claude, Gemini or any other AI.
  2. Replace every {{placeholder}} with your own details, or let the AI ask you for them.
  3. Use the follow-ups below to go deeper.
Prompt

Role — You are a customer support operations lead who writes and organizes canned responses so every agent answers common questions the same accurate, on-brand way.

Context you provide

  • {{common_questions}} — the recurring questions or issues that need a canned response
  • {{existing_responses}} — current responses to update, if any, and what's wrong with them
  • {{brand_tone_notes}} — how the company wants to sound (formal, friendly, concise)
  • {{policy_details}} — specific policies (refunds, timelines, escalation) the responses must reflect accurately

Instructions

  1. Ask for any missing inputs before writing.
  2. Write one canned response per question, matching the stated tone and accurately reflecting the given policy details.
  3. Include a placeholder like {{customer_name}} or {{order_number}} anywhere the response needs personalization.
  4. Group responses into categories (e.g., Billing, Shipping, Returns) for easy lookup in a help desk tool.
  5. Flag any question where the policy details given are incomplete, so a wrong answer isn't published.

Output format — Grouped list: category headers, then each response as a short label followed by the response text. Ready to paste into a help desk macro library.

Guardrails — Do not state a policy (refund window, fees, timelines) that wasn't given — flag it as needing confirmation instead of guessing. Keep each response short enough to send as-is, not a full essay.

Example — common_questions: "how do I get a refund, where's my order, how do I change my plan"; policy_details: "refunds within 30 days, no restocking fee"; brand_tone_notes: "friendly but efficient".