18 Copy-Ready AI Customer Support Prompts for Your Team

18 copy-ready prompts to analyze performance, audit conversations, write help articles and test your AI agent.

18 Copy-Ready AI Customer Support Prompts for Your Team

A vague prompt returns a confident answer you cannot verify. A good support prompt sets the scope, demands evidence and separates facts from guesses.

This library collects 18 copy-ready prompts from the Crisp 2027 AI Support Improvement Playbook. Each one targets a specific job: spotting a trend, finding the root cause of an escalation, drafting a help article, or testing a fix before it reaches customers.

TL;DR

  • 18 prompts across Copilot Analytics, Claude via Crisp MCP, Copilot Knowledge Base, Instructions and Playground.
  • The analysis prompts fix a period, a scope and an output format.
  • The analysis and writing prompts tell the AI to flag gaps instead of inventing facts.
  • Investigation prompts stay read-only. Changes happen after verification.
  • Never paste sensitive customer data you are not authorized to share.

The anatomy of a reliable support prompt

A reliable support prompt scopes the request, demands evidence, separates facts from hypotheses, bans invention and stays read-only. The analysis and writing prompts below follow these five rules. Apply them to any prompt you write yourself.

  1. Scope it. Name the period, the topic and the sample. "The last 30 days of billing conversations" beats "our conversations".
  2. Demand evidence. Ask for conversation references, supporting figures and the sample size.
  3. Split facts from hypotheses. Metrics show what changed. They never prove why.
  4. Ban invention. Tell the AI to mark missing facts, policies or links instead of filling the gap.
  5. Lock it read-only. Exploratory prompts say "do not change any data". Edits come after review.

All 18 prompts at a glance

PromptWhere to run itFramework step
Find your biggest AI improvement opportunityCopilot AnalyticsMeasure
Visualize an escalation trendCopilot AnalyticsMeasure
Build a weekly AI performance viewCopilot AnalyticsMeasure
Investigate failed renewal paymentsCrisp MCP + ClaudeInvestigate
Find what customers mean by “Where is my order?”Crisp MCP + ClaudeInvestigate
Discover questions missing from your FAQCrisp MCP + ClaudeInvestigate
Audit whether handovers are necessaryCrisp MCP + ClaudeInvestigate
Create a weekly support improvement briefCrisp MCP + ClaudeInvestigate
Understand onboarding frictionCrisp MCP + ClaudeInvestigate
Audit inconsistencies after a policy changeCrisp MCP + ClaudeInvestigate
Create a failed-payment help articleCopilot Knowledge BaseImprove
Rewrite an unclear returns articleCopilot Knowledge BaseImprove
Refresh an existing help article after a changeCopilot Knowledge BaseImprove
Make similar help articles easier to findCopilot Knowledge BaseImprove
Make AI behavior unambiguousInstructions / Review with AIImprove
Replay the conversation that exposed the problemPlaygroundTest
Test a policy answer and its exceptionsPlaygroundTest
Run the same weekly performance reviewCopilot AnalyticsRepeat

Copilot Analytics prompts: find what changed

Copilot Analytics prompts turn your Crisp Analytics data into a readout, a chart or a dashboard. The slash command sets the output type: /insights for a written readout, /chart for a visualization, /dashboard for a repeatable view.

1. Find your biggest AI improvement opportunity

Use it when: You have plenty of support data but do not know which issue to work on first. Example: A support lead sees more conversations reaching humans this month.

/insights
What changed in our support performance over the selected period compared with the previous comparable period? Highlight the most meaningful changes, their supporting figures, and three follow-up questions. Distinguish facts from hypotheses.

What you get: A short, evidence-backed shortlist of issues. Track: Autonomous resolution rate; escalation rate; conversation volume.

Guide: Copilot Analytics guide, Hugo Analytics guide

2. Visualize an escalation trend

Use it when: You need to see when an increase in escalations started. Example: Escalations rose after a product launch or a shipping campaign.

/chart
Create a chart to help me understand how AI escalations evolved during the selected period. If supported, include the rate and conversation volume over time. Label the metric clearly and explain any missing data.

What you get: A chart that helps narrow the timing for investigation. Track: Escalation rate and count.

Guide: Copilot Analytics guide, Hugo Analytics guide

3. Build a weekly AI performance view

Use it when: Your team needs a consistent view rather than a new report every week. Example: The support manager wants a dashboard for a Monday check-in.

/dashboard
Propose a weekly AI support performance dashboard using the available analytics. Prioritize autonomous resolution, escalation, response time, and conversation volume. Explain which metrics are unavailable and what each view helps us decide.

What you get: A repeatable monitoring view aligned with team decisions. Track: Resolution, escalation, response time, volume.

Guide: Copilot Analytics guide, Hugo Analytics guide

4. Run the same weekly performance review

Use it when: You want continuous improvement to become a team habit. Example: Your team reviews AI performance every Monday.

/insights
Compare the selected period to the previous comparable period. Summarize changes in relevant AI support outcomes and volume, note possible data limitations, and recommend three questions for our weekly support improvement review. Do not infer causes from metrics alone.

What you get: A consistent weekly review and a new evidence-based priority. Track: Baseline vs. current outcomes.

Guide: Copilot Analytics guide, Hugo Analytics guide

Claude prompts via Crisp MCP: investigate real conversations

Claude prompts via Crisp MCP investigate the conversations behind your metrics. Run them in Claude (or Mistral Le Chat) once you connect your workspace through Crisp MCP. Each one asks for conversation references and forbids changes.

5. Investigate failed renewal payments

Use it when: A recurring billing question often ends in a human handover. Example: “What happens if my renewal payment fails?”

Review the Crisp conversations you can access about failed renewal payments and subscription changes over the last 30 days. Group recurring questions, identify recorded human handovers, and propose the most likely root causes. Include conversation references, the sample size, and what should be verified in Observability. Do not change any data.

What you get: A ranked list of likely causes with supporting conversations. Track: Sampled handovers; repeated questions.

Guide: Crisp MCP setup guide, Observability guide

6. Find what customers mean by “Where is my order?”

Use it when: One broad topic contains different customer problems. Example: Customers ask about tracking, late orders, missing parcels or delivery estimates.

Review accessible Crisp conversations about shipping and delivery from the past 30 days. Group them into distinct customer intents, such as tracking, late delivery, or missing parcels. For each group, give representative questions, evidence references, observed handovers, and a recommended next investigation. State the sample limitations and avoid personal data.

What you get: A useful taxonomy for targeted knowledge or integration work. Track: Frequency within the reviewed sample.

Guide: Crisp MCP setup guide, Topics guide, Observability guide

7. Discover questions missing from your FAQ

Use it when: Your team suspects new customer questions are emerging. Example: A recent launch or policy change creates repeated questions.

Review accessible recent Crisp conversations and relevant knowledge articles. Identify recurring customer questions that appear unanswered, under-documented, or inconsistent with current articles. For each, include example phrasing, conversation references, the article if found, and a proposed documentation priority. Flag uncertain gaps; do not edit articles.

What you get: A candidate backlog for knowledge improvements. Track: Recurring question coverage.

Guide: Crisp MCP setup guide, AI Suggestions guide, Copilot Knowledge Base guide

8. Audit whether handovers are necessary

Use it when: You want to reduce avoidable handovers without removing necessary human oversight. Example: Routine questions trigger human intervention.

Analyze a sample of accessible Crisp conversations with recorded human handovers. Classify each as necessary, potentially avoidable, or unclear, explaining the reason. Cite conversation references and distinguish policy exceptions from knowledge or routing gaps. Report sample coverage; do not make changes.

What you get: A safer list of candidates for investigating escalation behavior. Track: Potentially avoidable handovers within sample.

Guide: Crisp MCP setup guide, Observability guide, Routing Rules guide

9. Create a weekly support improvement brief

Use it when: You need qualitative evidence that complements your Analytics dashboard. Example: A support lead wants a short readout for a weekly review.

Create a weekly improvement brief from Crisp conversations you can access in the last seven days. Include the reviewed sample size, top observed themes, notable customer friction, representative conversation references, suspected causes, three suggested actions and open questions. Separate observed evidence from interpretation; do not claim workspace-wide rates or modify anything.

What you get: A concise discussion document for the team. Track: Review coverage; validated action count.

Guide: Crisp MCP setup guide, Hugo Analytics guide, Copilot Analytics guide

10. Understand onboarding friction

Use it when: New users repeatedly ask what to do next. Example: Customers struggle to configure a workspace or invite their team.

Review accessible Crisp conversations about onboarding, setup, and invitations during the last 30 days. Identify recurring blockers by journey step, show representative anonymized questions and conversation references, and suggest whether each needs documentation, product feedback, or an agent investigation. Mark hypotheses and sample limits.

What you get: A grounded onboarding improvement backlog. Track: Recurring onboarding blockers.

Guide: Crisp MCP setup guide, AI Suggestions guide, Copilot Knowledge Base guide

11. Audit inconsistencies after a policy change

Use it when: Customer-facing policies have changed and answers may be inconsistent. Example: Return eligibility changed, but old wording still appears.

Using the official policy I provide below, review accessible Crisp conversations about returns or refunds since [date]. Flag responses that may conflict with the policy, cite the conversation references, and explain why. Separate confirmed discrepancies from ambiguous cases. Do not modify anything.

Approved policy: [paste policy]

What you get: A review queue for correcting inconsistent guidance. Track: Confirmed policy discrepancies in reviewed sample.

Guide: Crisp MCP setup guide, Copilot Knowledge Base guide

Copilot Knowledge Base prompts: write and fix articles

Copilot Knowledge Base prompts draft new help articles and fix existing ones inside the Crisp Knowledge Base editor. /create-article drafts from approved facts. /proofread rewrites an existing article and keeps its policies intact.

12. Create a failed-payment help article

Use it when: A repeated question has no reliable article to answer it. Example: Customers need steps for retrying a payment and keeping access.

/create-article
Draft a help article titled “What happens if my renewal payment fails?” Use only these approved facts: [paste policy and steps]. Cover what the customer sees, how to retry, what happens to access, and when to contact support. Use clear headings and short steps. Mark any missing facts instead of inventing them.

What you get: A review-ready article that answers the actual customer question. Track: Answer completeness; billing policy accuracy.

Guide: Copilot Knowledge Base guide, AI Suggestions guide

13. Rewrite an unclear returns article

Use it when: Customers misunderstand an existing return or exchange article. Example: Customers confuse the deadline for starting a return with the deadline for sending it back.

/proofread
Make the distinction between starting a return and sending it back unmistakable. Preserve all approved deadlines, fees, eligibility rules and exceptions. Use a clear example only if supported by this article. Flag contradictions or missing facts for review.

What you get: A less ambiguous customer-facing returns article. Track: Correct answers on policy edge cases.

Guide: Copilot Knowledge Base guide, AI Suggestions guide

14. Refresh an existing help article after a change

Use it when: A product, pricing or operational change made part of an article obsolete. Example: The steps to update a payment method or track a shipment have changed.

/proofread
Update this article using the following approved changes: [paste changes]. Remove obsolete instructions, keep accurate content, and clarify the customer journey. Do not invent steps, policies or links. List any questions I must resolve before publishing.

What you get: A maintained article without contradictory steps. Track: Stale instructions removed; answer accuracy.

Guide: Copilot Knowledge Base guide

15. Make similar help articles easier to find

Use it when: Customers use different wording for the same process and information is scattered. Example: “Cancel”, “unsubscribe” and “stop renewal” appear across overlapping articles.

/proofread
Improve this article’s title, introduction and headings so customers searching for [common phrases] can find the correct information. Preserve the original policies and links, remove redundant wording, and flag topics that belong in a separate article.

What you get: More precise, discoverable support knowledge. Track: Content coverage and discoverability.

Guide: Copilot Knowledge Base guide

Instruction example: make AI behavior unambiguous

This instruction example shows how to write one clear behavior rule for Hugo. Adapt it to your policy, then run Review with AI to catch conflicts.

16. Make AI behavior unambiguous

Use it when: The knowledge is accurate, but Hugo responds inconsistently. Example: Hugo sometimes asks unnecessary questions before giving an answer.

Instruction example (adapt, do not paste blindly):
When a customer asks a general policy question, answer using approved knowledge. Ask only for information required to resolve the issue. If the case requires approval or account-specific data you cannot access, explain the limitation and follow the appropriate escalation process.

What you get: A focused behavioral change without unnecessary new rules. Track: Consistent answers; correct handovers.

Guide: Instructions / Review with AI guide, Playground guide

Playground test inputs: prove the fix works

Playground test inputs prove a fix works before customers see it. Paste them into Playground after every change and test the original wording, a paraphrase and an exception.

17. Replay the conversation that exposed the problem

Use it when: You applied a targeted knowledge or behavior change. Example: The original question received a weak answer or unnecessary handover.

Example test prompts:
What happens if my renewal payment fails?
My card was declined, can I retry it?
I was charged twice and want a refund.

Replace with the actual question and edge case for your issue.

What you get: A reproducible test set with an expected result for each case. Track: Passed tests; appropriate escalations.

Guide: Playground guide

18. Test a policy answer and its exceptions

Use it when: You changed a returns article or routing rule. Example: A customer may ask about a normal return or an ineligible exception.

Example test prompts:
How long do I have to return my order?
Can I send my item back after the deadline?
My refund is missing and I want someone to investigate.

What you get: A regression check across normal and exception requests. Track: Passed policy tests; correct handovers.

Guide: Playground guide

Where each prompt fits in the improvement loop

Each prompt feeds one step of a five-step loop that turns customer conversations into better AI support. The prompt is never the goal. The fix is.

  1. Measure. Find where AI performance slips. Output: A prioritized problem backed by data.
  2. Investigate. Find the root cause in conversations. Output: A specific cause, supported by evidence.
  3. Improve. Fix knowledge or behavior. Output: One targeted, reviewed change.
  4. Test. Check the change before rolling it out. Output: A passed test with edge cases.
  5. Repeat. Monitor the effect and start again. Output: A measured result and next priority.

Use Analytics prompts to measure, Claude prompts to investigate, Knowledge Base and Instruction prompts to improve, Playground inputs to test, and the weekly /insights prompt to repeat.

Keep building the habit

This article belongs to the Crisp 2027 AI Support Improvement Playbook. Pair it with:

footer-cta-backgroundfooter-cta-patternfooter-cta-receptionfooter-cta-persona
  • footer-cta-badge-g2-high-performer
  • footer-cta-badge-g2-momentum-leader
  • footer-cta-badge-g2-loved

Ready to build exceptional customer support?