Prompt Library

ChatGPT support prompts that stay inside your policy, not the internet's nicest refund

15 copy-paste prompts

Paste ticket text, macro library, and policy excerpts you are allowed to cite. ChatGPT drafts replies and flags gaps as [VERIFY]. Agents approve before send.

In short: This page contains 15 copy-paste ready prompts, organized into 5 categories with a description and pro tip for each. The first 5 prompts are free instantly, no signup needed. Hand-curated and tested by the AI Academy team.

By Louis Corneloup · Founder, Techpresso

Last updated September 15, 2026·Hand-curated & tested by the AI Academy team

First reply

3 prompts

Acknowledge and triage

1/15

✨ What it does

ChatGPT drafts a first reply from ticket and policy with SLA placeholder. Agent confirms [VERIFY] SLA against the queue rules before send.

Customer ticket: [PASTE]. Policy excerpts: [PASTE POLICY]. Draft first reply: acknowledge issue, restate problem in one line, next step, timeline [SLA OR VERIFY]. Tone: [TONE]. No refund promise unless policy allows.

💡

Pro tip: Restate the problem in the customer's words. They need to feel heard before the fix.

Missing info request

2/15

✨ What it does

ChatGPT asks for [INFO LIST] in under 100 words with a why per item. Send once; bundle asks instead of three pings.

Ticket: [PASTE]. We need [INFO LIST] to proceed. Draft a short reply asking for those items with why we need each. Friendly, under 100 words.

💡

Pro tip: One email with three numbered asks beats three separate follow-ups.

Set expectations on wait

3/15

✨ What it does

ChatGPT drafts a wait-setting reply with ETA and workaround from your paste. Update ETA in the ticket if engineering slips again.

Ticket: [PASTE]. Known delay: [REASON AND DURATION OR VERIFY]. Draft reply with empathy, cause if shareable, new ETA, optional workaround from [PASTE WORKAROUNDS].

💡

Pro tip: Always give the next update date. Open loops churn CSAT.

Prompts get you started. Tutorials level you up.

A growing library of 300+ hands-on AI tutorials. New tutorials added every week.

Start 7-Day Free Trial

Macros from policy

3 prompts

Macro library row

4/15

✨ What it does

ChatGPT turns one policy rule into a macro with placeholders and when-not-to-use. Lead reviews [VERIFY] before the macro goes live in the tool.

Turn policy rule: [PASTE RULE] into a reusable macro: title, when to use, body with [PLACEHOLDERS], when not to use. Tag ambiguous cases [VERIFY].

💡

Pro tip: Every macro needs a when not to use line. That line saves escalations.

Personalize a macro

5/15

✨ What it does

ChatGPT personalizes the macro for the ticket while keeping policy sentences exact. Agent checks [ESCALATE IF] before clicking send.

Base macro: [PASTE]. Ticket: [PASTE]. Personalize names and specifics. Keep policy sentences exact. Flag if ticket needs escalation [ESCALATE IF].

💡

Pro tip: Do not paraphrase policy sentences in macros. Paraphrase is where liability lives.

Multi-language stub

6/15

✨ What it does

ChatGPT writes a translation stub noting fixed English terms and banned promises. Send stub to your localization vendor; do not auto-translate policy.

English macro: [PASTE]. Write a stub note for translation to [LANGUAGE]: terms that must stay English, placeholders, banned promises. Not a full translation.

💡

Pro tip: Legal terms often stay English. Flag them so translators do not improvise.

Escalation and de-escalation

3 prompts

Escalation note internal

7/15

✨ What it does

ChatGPT drafts an internal escalation note with tried steps and policy blocker. Tier-2 should answer within one SLA period.

Draft internal escalation note for ticket [ID]. Customer said: [PASTE]. Tried: [PASTE STEPS]. Policy blocker: [PASTE OR VERIFY]. Ask tier-2 for: [ASK]. Facts only.

💡

Pro tip: Escalations without tried steps get bounced. Paste what you already did.

De-escalation reply

8/15

✨ What it does

ChatGPT drafts a de-escalation reply offering only remedies from policy. Supervisor approves if remedy touches billing.

Angry ticket: [PASTE]. Allowed remedies from policy: [PASTE]. Draft calm reply: validate, fact check from ticket, offer allowed remedy, next step. No fight bait.

💡

Pro tip: Validate once, then move to the allowed remedy. Double validation sounds fake.

Supervisor approval ask

9/15

✨ What it does

ChatGPT drafts a supervisor approval note with policy cite placeholder. Wait for written yes before you promise [REMEDY].

Draft a short note to supervisor: customer [ID], issue [ISSUE], remedy I want to offer [REMEDY], policy cite [CITE OR VERIFY]. Request yes/no.

💡

Pro tip: Get approval in writing in the ticket. Verbal yes disappears on shift change.

Like these prompts? There are full tutorials behind them.

Learn the workflows, not just the prompts. 300+ easy-to-follow tutorials inside AI Academy — and growing every week.

Try AI Academy Free

Billing and refunds

3 prompts

Refund eligible

10/15

✨ What it does

ChatGPT drafts refund eligible or decline reply from policy with amount placeholder. Billing confirms [AMOUNT] before customer sees it.

Ticket: [PASTE]. Refund policy: [PASTE]. If eligible, draft reply with amount [AMOUNT OR VERIFY], method, timeline. If not, decline with allowed alternative.

💡

Pro tip: Never round refund amounts. Paste the invoice line if you have it.

Partial credit offer

11/15

✨ What it does

ChatGPT drafts a partial credit offer with value and expiry placeholders. Finance sets [VALUE]; agent sends after approval.

Policy allows partial credit: [PASTE]. Issue: [ISSUE]. Draft offer with credit value [VALUE OR VERIFY], expiry, how to apply. Customer-facing tone.

💡

Pro tip: Partial credit needs expiry and apply steps. Vague credit creates second tickets.

Billing dispute facts

12/15

✨ What it does

ChatGPT explains invoice line items in plain language from your facts. Resolve [VERIFY] items with billing before reply.

Customer disputes charge: [PASTE]. Invoice facts: [PASTE]. Draft reply explaining line items in plain language. Unknown items [VERIFY]. No legal threats.

💡

Pro tip: Quote invoice dates and SKUs. Vague billing replies feel like stonewalling.

QA and coaching

3 prompts

QA rubric score

13/15

✨ What it does

ChatGPT scores the reply against your rubric with three fixes. Coach uses fixes in a 10-minute session, not as auto-send.

Score this agent reply: [PASTE REPLY] against rubric: [PASTE RUBRIC]. Ticket context: [PASTE]. Output: score per dimension, 3 fixes, no rewrite yet.

💡

Pro tip: Score against rubric dimensions, not vibes. Agents need repeatable feedback.

Coaching note

14/15

✨ What it does

ChatGPT writes a 100-word coaching note with one fix and one macro idea. Manager delivers live; do not paste customer PII into ChatGPT on production accounts.

From ticket thread: [PASTE], write a 100-word coaching note for agent [NAME OR ROLE]: what went well, one fix, one macro to save.

💡

Pro tip: One fix per coaching note. Three fixes and the agent remembers none.

Knowledge gap flag

15/15

✨ What it does

ChatGPT lists KB gaps from three tickets and proposes one article outline. Content team owns [VERIFY] policy gaps with legal.

Tickets this week about [TOPIC]: [PASTE 3 TICKETS]. List knowledge base gaps and propose one help article outline. Tag policy gaps [VERIFY].

💡

Pro tip: One article outline beats a vague update the help center note.

Free tool

Prompt Optimizer

Turn a rough idea into a structured, professional AI prompt.

Try it free →

Frequently Asked Questions

No. These prompts stay inside ChatGPT's official product. It drafts inside pasted policy; agents approve before send. If a setting is missing on your plan, the prompt says so.

A current ChatGPT account. Some features sit on paid plans. The prompts still work as checklists if a feature is off.

Every prompt tells it to use only what you paste and to tag gaps as [VERIFY]. You still check the support reply before you save it.

Yes to copy. Running them needs your ChatGPT account. Long files fit better on a paid plan.

No. It is a draft of your support reply. You own the facts, the files, and whatever your job or policy requires.

Prompts are the starting line. Tutorials are the finish.

A growing library of 300+ hands-on tutorials on ChatGPT, Claude, Midjourney, and 50+ AI tools. New tutorials added every week.

Start 7-Day Free Trial Explore AI for Marketers

7-day free trial. Cancel anytime.

Browse All Prompt Collections

Popular prompt collections

AI Photo Editing Prompts ChatGPT Jailbreak Prompt Best ChatGPT Prompts Funny Hinge Prompts ChatGPT Caricature Prompts