The UXCam MCP goes furthest when your assistant combines UXCam with everything else it can reach: your other apps, your files, your code and your team. Below are ideas to get you started.
Each comes with a prompt you can copy, replacing the words in brackets with your own cases, or simply use them as inspiration.
Compare apps, regions and platforms side by side
In the dashboard you look at one app at a time. Through the MCP, your assistant can query several apps in one go and put them in a single report. This is useful if you run separate apps per region, brand or platform.
Compare [Shop DE], [Shop UK] and [Shop US] for the [last 30 days]:
sessions, users, average session duration, crash rate and rage gesture rate.
Then compare conversion in the [checkout] funnel from [Cart] to [Order Confirmation].
Show the results in one table and tell me which app is underperforming and where.
Connect UXCam with your own data
UXCam tells you how people use your app. Your users' data tells you who they are and what they're worth. Connect them to get detailed user profiles, understand their behavior, analyze revenue, create marketing campaigns, etc. Attach a file or connect your other tools (CRM, Data warehouse, internal docs, help center, etc).
Revenue: find out what your most valuable users do differently.
Using our [Stripe] data, find the top [10%] of users by spend in the [last 30 days].
Then use UXCam data for [App name] to compare their behavior with everyone else:
how often they open the app, session length, the screens and features they use most,and where they hit friction.
What do our highest spenders do that others don't, and which of those behaviors could we encourage in other users?
Or look at the long term:
Group our users into [3] lifetime value tiers using [our data warehouse].
For each tier, show what their first [7 days] in [App name] looked like in UXCam:
onboarding completion, features used and sessions per week.
Which early behaviors best predict a high lifetime value?
Customer profiles: build behavioral profiles for your segments.
Using our [CRM] data for accounts on the [Enterprise] plan and UXCam data for [App name], create a behavioral profile for each segment: how often they use the app, which screens and features they rely on, and where they struggle.
Present it as a one-page summary per segment.
Research: check what users say against what they do.
Here are the notes from our [5 onboarding interviews] (attached). For each pain point mentioned, check UXCam data for [App name] from the [last 60 days]: how many users hit that step, how many drop off there, and show me sessions that illustrate it.
Which findings are confirmed by the data and which aren't?
Give your design agent real user evidence
Connect UXCam to the assistant or agent your designers use, so design changes start from how users actually behave. It can pull heatmaps, drop-off and session evidence for a screen, then propose changes to the design.
Look at the [Checkout] screen in [App name] for the [last 30 days]: the heatmap,
rage taps, unresponsive gestures and sessions where users left without continuing.
Summarize the top 3 usability problems with evidence, then propose design changes
for our [Checkout design in Figma] that address each one.
Turn frustration into code fixes
In a coding assistant such as Cursor or Claude Code, your assistant can find where users struggle and then look for the cause in your codebase.
Find the screens in [App name] with the highest UI freeze and rage gesture rates
in the [last 14 days], on app version [4.2.0]. For the worst one, find the code
that renders that screen in this repository and suggest what could be causing it.
Check the impact of a release
Compare a new version with the one before, before your users tell you something broke.
Compare app version [4.2.0] with [4.1.0] in [App name] for sessions since [release date]:
crash rate, UI freeze rate, app start time, session duration and conversion in the
[onboarding] funnel.
Flag anything that got worse and show me example sessions.
Explain a change in your numbers
When a metric moves, ask your assistant to break it down instead of digging through reports.
[Sign-ups] in [App name] dropped [last week] compared to the week before.
Break the change down by platform, app version, country and device.
Check whether we lost any sessions to data collection issues, then show me sessions from the segment with the biggest drop. What is the most likely cause?
Write specs and tickets backed by data
Bring real numbers into product docs. Connect your assistant to your docs or issue tracker and it can draft the document or ticket with UXCam evidence included.
We're planning to redesign [onboarding] in [App name]. Pull the current funnel conversion, the biggest drop-off step, retention of users who complete it versus those who don't, and 3 example sessions.
Draft a problem statement and success metrics for the spec in [Confluence].
Investigate a user's issue
When a customer reports a problem, see what actually happened before you reply.
User [user ID] reported [that the payment screen froze]. Find their sessions in
[App name] from the [last 7 days], describe what happened in each, and tell me whether other users hit the same issue on the same screen and app version.
Set up a weekly routine
If your assistant supports scheduled or recurring tasks, have it send you a summary automatically. Otherwise, save the prompt and run it every Monday.
Every [Monday at 9:00], create a weekly summary for [App name] covering the [last 7 days] compared with the week before: sessions, users, crash rate, rage gesture rate, [checkout] funnel conversion and the screens with the most frustration.
Highlight the 3 biggest changes and include example sessions. Send it to [#product in Slack].
Keep an eye on privacy
Check regularly that no personal data is showing up in your recordings.
Check [App name] for sessions in the [last 30 days] where personal data was detected on screen.
List the screens and types of data involved, and tell me which screens need masking.
Give your whole team easy answers
Not everyone on your team knows their way around analytics. Anyone with UXCam access can connect their own assistant and ask questions in plain language. You can also set up a shared project in your team's assistant with a few instructions about your app, so everyone gets consistent answers.
You help our team answer questions about [App name] using UXCam.
Our key flows are [onboarding], [search] and [checkout]. Our main screens are called [...].
Always state the date range and filters you used, and link to example sessions where useful.
Then anyone can ask:
How many people used [the new search filters] last week, and did they find what they were looking for?