The Best User Research Skills for Claude Code

Most "user research" done by solo builders is not research. It is a conversation that confirms what they already believed. These skills are mostly about interrupting that.

The user-research cluster in our directory is unusual: it includes skills that interview you as well as skills that structure interviews with your users. Both are useful, and they solve different problems.

The skills worth installing

interview-me — 9.4 · Developer Tools

Extracts what you actually want instead of what you think you should want. This is the highest-scoring skill in the group and the least obviously "research" — but the biggest source of bad product decisions is an unexamined premise, not a missing survey.

linkedin-interviewer — 9.4 · Productivity

Interviews you for the raw material behind your posts — the experiences and numbers you have not written down. Filed under LinkedIn, but the technique is general: structured interviewing of yourself is a research method.

founder-consumer — 9.3 · Developer Tools

Spins up a swarm of buyer sub-agents (100 by default) to act as a simulated consumer panel. Treat this as an idea generator, not evidence — simulated buyers reflect the assumptions in the prompt. It is useful for surfacing objections you had not considered, and dangerous if you cite it as validation.

founder-marketing — 9.1 · Productivity

Turns the simulated panel and a competitor map into positioning. The natural downstream of founder-consumer.

founder-offer — 9.0 · Productivity

Builds an offer by enumerating every problem the buyer hits on the way to the outcome. The offer is usually the real product decision, and it is usually made by accident.

founder-pricing — 8.8 · Developer Tools

Finds the price buyers will actually pay, using the panel's responses. Same caveat as above: simulated responses are hypotheses, not data.

interview-script — 8.1 · Developer Tools

Structured customer interview scripts with Jobs-to-be-Done probing questions. The most conventional research skill here, and the one to use when you are actually talking to real people.

summarize-interview — 7.9 · Writing

Turns an interview transcript into a structured template with JTBD framing. The complement to interview-script — one produces the conversation, the other files it.

Real users vs simulated ones

The single most important distinction in this list is between skills that work with real input and skills that simulate it. interview-script, summarize-interview, and interview-me operate on things that actually happened. founder-consumer, founder-pricing, and to a lesser extent founder-marketing generate plausible input.

Simulation is useful for expanding the space of hypotheses. It is not evidence, and the scores here do not change that.

How to choose

• Talking to real users? interview-script + summarize-interview.
• Clarifying your own premise? interview-me.
• Exploring positioning? founder-consumer → founder-offer → founder-marketing, keeping in mind what the inputs are.

Every skill page carries the written rationale behind its score; methodology explains what the dimensions measure.