These are the Claude Skills, research agents and prompts I use on my own studies every week, built from my real research so Claude follows a solid process and the first draft is one you can build on
One-time payment and lifetime access, instead of $315 if you bought them separately
I walk through some of the agents, skills and prompts in the bundle, so you can see what they look like and how they work before you buy
Leadership keeps saying "do more, go faster, use AI," and nobody has shown you what that looks like on a real study
You asked Claude for a research plan twice, the same way, and got two completely different plans back
The AI synthesis comes back flat and hollow, and when you check it against your transcripts, some of it isn't there at all
You spend so long editing and double-checking the output that you wonder whether it saved you any time
Stakeholders send you AI-written briefs that sound confident and fall apart the moment you ask a follow-up question
You're worried that if AI does the first pass, you'll stop feeling like you own your own research
I hear some version of these on almost every mentorship and training call I run, and I recognize every single one, since I've been there myself
I've been a user researcher for 13 years and using AI in my research for over three. When I started, I'd paste in a transcript, ask for themes, and get back something so flat and hollow that I'd sit there thinking, why didn't I just do this myself? Did it really save me any time?
So I tried to fix it by writing a research plan skill that was supposed to make great plans. And, guess what, it didn't, because I hadn't given it any context about how I plan research, what a good plan looks like to me, or the decisions my plans need to support.
Things started to change once I built every skill from my previous work, using the discussion guides, synthesis and reports I already trusted. Claude followed my process instead of guessing at one, and I stopped re-explaining myself at the start of every study.
Now I still do my own affinity diagramming, and I have Claude run synthesis alongside me so I can compare. More than once I've looked at its version and thought, "oh yeah, I completely missed that."
AI output is always a first draft for me, never the final draft, and the judgment stays mine. What changed is that the first draft is finally worth building on.
When Claude doesn't know your process, it fills the gaps with generic answers, and you end up writing the same long prompt on every study. Here's what I wanted instead, and what I hear researchers ask for on calls:
Output that follows a solid research process every time, so two people on your team asking for the same thing get the same quality back
The checklist, mundane work off your plate, so you have more time for stakeholders and the strategic stuff
Your interviews, survey and support tickets pulled into one holistic view, instead of scattered across a dozen tabs
A sparring partner that pokes holes in your plan before your stakeholders get the chance
The judgment and the final call staying with you, with AI handling the labor around the research
A real answer, with examples from your own studies, the next time someone asks how you're using AI
It's my Claude Skills, my research agents and my AI Prompt Library together, so you get to skip the trial and error I went through and start using AI on your real research this week
76 skills I built from my own interview guides, synthesis and reports, so you convey your expertise one time and Claude follows a solid research process whenever you ask it for something.
Nine plugins, one for each stage of a research project. A skill is one job, and an agent is a coordinator that runs several steps toward a goal, calling in the right skills and connectors at each one.
109 prompts for every stage of research, from study design and recruiting through interviews, synthesis and readouts. They work in Claude, ChatGPT, Gemini or Copilot, so anyone on your team can use them.
I think of these as three levels, and I reach for a different one depending on how much of the work I want AI to take on for a particular task
When I need a recruitment email, a workshop agenda, or a second opinion on my research questions, I grab a prompt from the library and paste it into whatever AI tool I have open
For anything I do on every study, like a discussion guide or a topline, I use a skill so Claude follows the same process each time and I don't have to explain it again
When I want a full stage handled, like turning ten transcripts into themes and insights, I hand it to an agent, which pulls in the right skills and connectors on its own
"There is nothing out there with the quality of your tools and content. This bundle has saved me so much time and energy right from the start and I know it's something I can trust"
"I got your AI bundle a month ago and, since then, it's been so incredibly helpful in making sure I stay consistent, always connect my research to impact, and also has allowed me to enable stakeholders to actually write good research plans and screeners! These resources are an absolute gem!"
"Honestly, the sparring agents are amazing. It feels like I'm having a conversation with you and it has helped me immensely with brainstorming and not feeling so alone in my role. Thank you!!"
"Thank you so much for continuing to update the skills, agents, and prompts. There is so much noise out there and these allow me to cut through that to quality resources that actually WORK. I use them in my workflows every single day and they've not only saved me time but also my sanity of the sheer months it would take me to create these on my own."
You start from a draft that already follows your process, so your time goes into checking and sharpening it instead of writing from a blank page
Your team gets the same quality back no matter who runs the study, since everyone is working from the same skills and the same standards
You get time back for the part of the job that gets research used, like stakeholder conversations, readouts and pushing on decisions
| Figuring it out on your own | The UXR AI Bundle | |
|---|---|---|
| Context | ×Writing long prompts from scratch and re-explaining your process on every study | ✓Skills that already hold a solid research process, built from real studies |
| Consistency | ×Asking the same question twice and getting two different answers | ✓The same process every time, for everyone on your team |
| Trust | ×Copying prompts from strangers without knowing what's inside them | ✓Plain text files you can open and read before you use them |
| Time | ×Months of trial and error, which is what it took me | ✓Installed and running on your next study |
| Coverage | ×A prompt here and there for one stage of a project | ✓Prompts, skills and agents for every stage, from intake to reporting |
P.S. Need an invoice to expense this? Make the purchase and then email me with your org's details and I will create a custom invoice for you!
The prompts work in any AI tool, including the free versions of Claude, ChatGPT and Gemini. The skills need Claude, and the agents run in Claude Cowork, which needs a paid Claude plan.
I don't fully trust it either, so I treat everything it gives me as a first draft. I still do my own affinity diagramming and compare it with what Claude comes back with, and I recommend the same habit to you, since that comparison is how you catch anything that's missing or doesn't match your data.
I handle sensitive data in my own consultancy, so the guides cover how I use AI with research data, including what I keep out of AI tools and how I summarize before sharing anything. Your company's AI policy comes first, and I recommend checking it before you upload participant data anywhere.
Every skill is a plain text file, so you can open it and read exactly what it tells Claude to do before you install anything. I recommend doing that with any skill you download, mine included.
Yes! The prompts work in any AI tool, so anyone on the team can start with those, and the skills and agents are there for whoever has Claude.
No. Each product comes with setup instructions, and if you can upload a file to Claude you can install the skills and plugins. The prompts are copy and paste.
Yes! Email me with what you already have and I'll sort out a price for the other two so you aren't paying twice.
Since these are digital downloads, I can't issue full refunds, but if you are not satisfied, please email me and we can work something out.
I recommend picking one study you have coming up, running one skill on it, and seeing how the first draft compares to what you'd normally write. Eventually, you'll build a rhythm that's all your own!
Get all three for $249