TL;DR: A practical prompt library for using Userflow MCP. 50 real prompts, organized by job-to-be-done, that you can copy into Claude or ChatGPT—plus tips for getting the most out of each one.

Key Takeaways

• Userflow MCP connects your Userflow data to Claude and ChatGPT
• Prompts in this library cover seven use cases: daily performance, onboarding diagnostics, NPS analysis, account lookups, segment analysis, documentation queries, and account audits
• Specific prompts outperform vague ones; include flow names, segment names, and time ranges

How to Use These Prompts

To put these prompts to work, connect Userflow MCP to your AI tool of choice. Setup is secure and based on each user’s permissioning on Userflow.

This article is a practical prompt library for using Userflow MCP. The prompts below are grouped by job: reviewing performance, diagnosing onboarding, summarizing NPS feedback, looking up specific accounts, and more. Copy them directly into Claude or ChatGPT—or adapt them to your own flows and segments.

5 Tips for Getting the Most out of Userflow MCP

Five things that make every prompt work better.

1. Be specific about flow names, segments, and time ranges. "Show me the Welcome Tour completion rate for the last 30 days" works better than "what's the welcome tour rate?" The more concrete your reference points, the more accurate the answer.
2. Iterate in the same conversation. Start broad, then drill down. After a summary, follow up with "show me the actual quotes" or "break this down by plan tier" to dig into specifics.
3. Reference Userflow explicitly when you've connected multiple data sources. "From Userflow, what's the trend of Flow Started events…" prevents your AI from confusing Userflow data with other connected MCPs.
4. Save prompts you run regularly. Set up a Claude Project, Custom GPT, or saved prompt for your most common queries. Turn the morning dashboard check into a one-prompt routine.
5. Combine with other MCPs. If you've connected Userflow MCP alongside HubSpot, Slack, or Mixpanel MCP, ask cross-tool questions in a single prompt: "Show me Userflow flow-completion rates for accounts in HubSpot's at-risk segment."

Your Daily Performance Check Prompts

For PMs, growth teams, and anyone who used to start their day with a Userflow dashboard. Ask once, get the answer, move on.

Example prompts:

1. "How many users saw [Flow Name] yesterday?"
2. "What's the completion rate for [Flow Name] this month?"
3. "Has [Flow Name] seen more views today than yesterday?"
4. "Show me the trend of Flow Started events for the last 7 days."
5. "How many users gave a positive NPS in the last 30 days?"
6. "What's the average time to complete [Flow Name]?"
7. "Give me a summary of [Checklist Name] performance for Enterprise users."

Pro tip: Save common prompts as templates in your AI tool. Many teams set up a "morning check" in Claude Projects or Custom GPTs that runs the same questions every day.

Ready to run these prompts? Connect Userflow MCP in minutes. Start a free trial or see the setup guide.

Onboarding and Product Adoption Diagnostics, Without the Analyst

When something is breaking in onboarding and you need answers now—not next sprint.

Example prompts

8. "Which step in [Flow Name] has the highest drop-off rate?"

9. "Where are users dropping off in our signup-to-activation flow?"

10. "Compare completion rates of [Onboarding A] and [Onboarding B]."

11. "Does [Flow X] have a higher drop-off at Step 2 compared to [Flow Y]?"

12. "Show me a side-by-side view of engagement for the two trial-ending flows."

13. "Tell me which flow performs better for mobile users specifically."

14. "Which flow is driving more Feature Adoption events: A or B?"

Pro tip: Compare two flow versions before you ship. Run a single prompt comparing the existing flow to the new variant in your AI tool—no chart-building required.

Prompts to Turn NPS into Clear & Actionable Themes

For CS, product, and execs prepping for a Board meeting or QBR. No need to read every response—ask for the patterns.

Example prompts

15. "Summarize the NPS text responses submitted in the last 30 days."

16. "What percentage of users responded to the NPS survey in [Feedback Flow]?"

17. "Compare NPS scores between [Old Feedback] and [New Feedback] flows."

18. "How many customers gave positive NPS in the last 30 days?"

19. "Show me the trend in NPS scores over the last quarter."

20. "What are the most common themes in NPS detractor feedback?"

21. "List the NPS responders who mentioned [Feature Name]."

Pro tip: Ask follow-up questions in the same conversation. After getting a summary, ask "show me the actual quotes" or "group these by theme" to drill in further.

Look Up a User or an Account, Fast

For CSMs prepping for a renewal call, sales answering "have they actually used it yet?" questions, or support diagnosing a specific user issue.

Example prompts

22. "Did [user@email.com] complete the [Setup] flow?"

23. "Has [Acme Corp] triggered the Plan Upgraded event yet?"

24. "When was the last time [User Name] saw the launcher?"

25. "What was the last event tracked for [Company Name]?"

26. "Did anyone from [Company] see the [Announcement Name]?"

27. "Is [Acme Corp] currently seeing any active flows?"

28. "Has [user] ever clicked the Invite Team button?"

‍Pro tip: Pull these up two minutes before a renewal call. CSMs can ask their AI for an account summary covering recent flow activity, NPS, and any flagged events.

Segment and Funnel Analysis to Make the Right Product Adoption Decisions

For PMs, growth, and analysts pulling lists, comparing populations, and tracking cohort behavior.

Example prompts

29. "How many users are currently in the [Churn Risk] segment?"

30. "List all companies that belong to the [Top Tier] segment."

31. "What percentage of our total user base is in [New Signups]?"

32. "Which users entered the [Inactive] segment in the last 24 hours?"

33. "How many users on the Free plan have performed the Export event?"

34. "Count users who completed Onboarding but haven't logged in for 10 days."

35. "Show me the top 10 companies by Flow Completion count."

36. "Give me a list of emails for users who haven't started any flow yet."

‍Pro tip: Use these for board prep. "How has the New Signups segment grown over the last quarter?" is a one-prompt answer to a question that used to need a separate dashboard.

Get Answers from Userflow's Docs Without Tab-Switching

For new admins, builders learning a new feature, or anyone who's ever had to leave their AI tool to find a help article.

Example prompts

37. "How do I trigger a flow based on a custom attribute?"

38. "What's the difference between a flow and a launcher?"

39. "How do I set up the Userflow API?"

40. "How do I use FlowAI Builder?"

41. "What's the best way to onboard mobile users specifically?"

42. "How do I export a list of users who completed a checklist?"

43. "Walk me through setting up a Resource Center."

Pro tip: Answers come with citations linking back to Userflow's documentation. If your team is new to Userflow, this replaces the "go read the help center" step in onboarding.

Spring Cleaning for Your Userflow Account

For admins managing accounts with hundreds of flows. Find what's stale, who owns what, and what to retire.

Example prompts

44. "Show me a list of all unpublished flows that haven't been edited in 3 months."

45. "Which flows are currently archived but were active in January?"

46. "List all flows tagged with [Marketing]."

47. "Who was the last person to edit [Flow Name]?"

48. "How many flows are currently in Draft status?"

49. "List all flows created by [Mark] in the last 30 days."

50. "What's the total count of active flows in production?"

‍Pro tip: Run this quarterly. A "flow audit" prompt before each planning cycle saves 30 minutes of dashboard hunting.

Frequently Asked Questions

What is Userflow MCP? Userflow MCP is a Model Context Protocol server that connects Userflow data to AI tools like Claude, ChatGPT,and Claude Code. Once connected, you can query adoption insights, flow completions, NPS responses, funnel performance, segment membership, and user behavior in plain language, with no SQL or dashboard required.

Which AI tools support Userflow MCP? Userflow MCP works with Claude, ChatGPT, and Claude Code.

‍How long does Userflow MCP setup take? Setup is seamless and secure. Connect your Userflow account to your AI tool using the MCP configuration. Userflow's setup guide walks through the steps.

Can I use Userflow MCP prompts with Cursor or Claude Code? Yes. Software builders and engineers can connect Userflow MCP inside or Claude Code and query product data during development. This is useful for incident response, debugging flow triggers, and checking recent error events without leaving the editor.

Do I need SQL to use Userflow MCP? No. All queries use plain language. You type a question in your AI tool, and Userflow MCP returns the answer from your Userflow data. No SQL, no chart setup, and no CSV export required.

Can I combine Userflow MCP with other MCPs? Yes. If you have connected Userflow MCP alongside HubSpot, Slack, or Mixpanel, you can ask cross-tool questions in a single prompt. For example: "Show me flow completion rates for accounts currently in HubSpot's at-risk segment."

Are these prompts included in Userflow's free plan? Userflow MCP is available on every Userflow plan. The prompts in this library work regardless of which plan you are on.

How do I save prompts I run every day? Set up a Claude Project, Custom GPT, or saved prompt in your AI tool for your most common queries. Many teams build a "morning check" routine with the daily performance prompts in this library.

Can I use Userflow MCP to query NPS feedback? Yes. You can ask your AI to summarize NPS responses, identify common themes in detractor feedback, list users who mentioned a specific feature, or compare NPS scores across two flows.

Will more prompts be added to this library? Yes. Userflow will expand this library as new use cases emerge and as we add support for native AI marketplaces.

Is Userflow MCP secure, and who can access my data? Yes. Userflow MCP connects through a secure MCP configuration, and access is governed by each user's existing Userflow permissions. Connecting an AI tool doesn't grant access to anything a user couldn't already see in Userflow—each connection is scoped to that user's role, so it can't expand or bypass existing permissions.

What is Userflow MCP?

Userflow MCP is the bridge between Userflow and the AI tools your team already uses. It works in Claude, ChatGP, and Claude Code today via custom MCP, with native marketplace availability coming next.

These prompts are one way to start. The library will keep growing as new use cases emerge and as we add support for native AI marketplaces.

Explore Userflow MCP →