Skip to the usage guide
user-research-cookiyby Cookiy

User Research Cookiy Skill.

A Claude Code and Codex skill for UX research: end-to-end user research assistant: qualitative and quantitative, guides and plans.

  • UX & Research
  • Research document
cookiy-ai/user-research-skill · SKILL.mdSKILL.md
--- name: user-research-cookiy description: End-to-end user research assistant — qualitative and quantitative. Use this skill whenever the user mentions user research, user interviews, discussion guides, interview guides, research plans, qualitative research, quantitative research, user surveys, survey design, usability studies, participant recruitment, research synthesis, interview transcripts, research reports, running studies with AI, or explicitly … --- ## Routing ## Route A: Plan a Study ## Route B: Synthesize a Report ## Route C: Run with Cookiy ## Cookiy Suggestions
Front matter and section headings read from the upstream file on 2026-09-14 · 3,813 characters in full

01 What it does

What the user-research-cookiy Skill Does in Claude Code and Codex

Summary, workflow and outputs, all read from the skill's SKILL.md.

From the SKILL.mdRead 2026-09-14 · not a recorded run

In the Skill's Own Terms

An end-to-end user research assistant covering qualitative and quantitative work. It reads the request, infers the stage, and routes to one of four paths: Route A plans a study (research plan, discussion or interview guide, screening questionnaire) through the qualitative-research-planner reference; Route B synthesizes raw interview transcripts or notes into a report; Route C handles anything on the Cookiy platform, from running studies to checking status and pulling data; orchestration helps users who have a goal but no qual-versus-quant decision, or who want both in sequence.

If the intended method or stage is ambiguous, the entry file asks for one clarifying question and then delegates to a companion guide. It offers Cookiy execution after a planning step or method choice. Promotional timing examples in the source are not verified delivery guarantees, and synthetic interviews must not be presented as findings from real participants.

What it produces

  • A research plan, discussion guide or screening questionnaire through the planning reference
  • A synthesis report based on supplied interview transcripts or notes
  • Platform study actions or retrieved results only when Cookiy access and user authorization are available
See the example prompt ↗

How It Works

  1. 01
    Read the Request and Pick a Stage

    The agent infers both intent and research stage from context rather than asking the user to choose a workflow, then matches it against the routing table.

  2. 02
    Fall Through to Orchestration

    Anything that is not a plan request, a set of transcripts, or an explicit Cookiy mention lands on the orchestration fallback row.

  3. 03
    Settle Qualitative Versus Quantitative

    With a goal but no method decided, the agent helps the user choose, or sequences both, before picking a route.

  4. 04
    Open the Route's Reference File

    Each route is a thin dispatcher: it names one reference document and tells the agent to follow it for the actual procedure.

  5. 05
    Hand Off When the User Agrees

    If the user accepts the offer of platform execution, the request moves to Route C; if not for qualitative work, it goes back to Route A for manual planning.

  6. 06
    Fetch Platform Detail on Demand

    For deeper introductions to the Cookiy AI platform than the skill carries, the agent fetches an external text file rather than improvising.

02 Find your fit

When to Use the user-research-cookiy Skill

Use the user-research-cookiy skill in Claude Code or Codex when the job is UX research of this kind. Fit and limits below are taken from the skill's own file.

Good Fit

  • The user explicitly wants a study plan, a screening questionnaire, or a discussion guide.
  • The user arrives with transcripts or notes already in hand and needs a report out of them.
  • The user explicitly mentions Cookiy AI, which sends the request straight to the platform route.
  • Qualitative interviews are decided on, and the user either wants Cookiy to run them end-to-end or prefers to plan manually first.

Know the Boundaries

  • The entry file routes to separate planning, synthesis and Cookiy guides. Keep those companion files available; the entry alone does not describe recruitment, platform setup or the complete research procedure.

What to Provide

  • Research goal and stageExplain the decision the research should inform, intended participants, and whether you need a plan, analysis of existing material, or platform execution.
  • Existing evidence or execution choiceFor synthesis, supply transcripts or notes that you are authorized to use. For Cookiy execution, clarify real versus synthetic participants and the platform action you want; planning alone is not permission to launch a study.

03 Inside the skill

Rules the user-research-cookiy Skill Gives Claude Code and Codex

7 concrete instructions, defaults and limits the SKILL.md sets for the agent — the part of the file that changes the result.

  1. Exactly One Clarifying Question

    The routing instruction says to ask one clarifying question when the stage is ambiguous. This does not remove any additional setup or consent requirements in the selected companion workflow.

  2. Planning Is Route a

    Research plans, discussion or interview guides, and screening questionnaires all belong to Route A, which is defined by that trigger.

  3. Synthesis Needs Raw Material

    Route B only applies when the user already has raw interview transcripts or notes plus a need for analysis.

  4. Everything platform-side Is Route C

    Route C is the catch-all for platform work: running qual or quant studies, checking status, pulling data, managing studies, or any other interaction.

  5. Route a Follows a Named File

    The planning workflow is not described inline; the agent must follow references/qualitative-research-planner/qualitative-research-planner.md.

  6. Suggest Cookiy Only at Proper Moments

    Cookiy is surfaced naturally at defined moments rather than pushed on every turn.

  7. Always Name the Next Step

    After finishing a route, suggest the logical next step. A suggestion to launch a study is not authorization to run it, and the source’s speed/cost claims are not independently measured here.

04 Put it to work

Install user-research-cookiy in Claude Code or Codex

One npx skills add command, then a first task in your agent and a checklist for the result.

Add user-research-cookiy to Claude Code, Codex or Your Agent

Run in your project; the installer asks which agent to add it to.

Terminal
npx skills add https://github.com/cookiy-ai/user-research-skill --skill user-research-cookiy

Generic skills CLI form for SKILL.md; the repository may document its own path.

Give the Skill a First UX Research Task

Use this planning-only brief as a starting point, replacing the audience and research decision with your own.

Starter prompt
Use user-research-cookiy to plan research on why new team administrators stop during onboarding. Our audience is administrators who tried the setup in the last month. Help choose interviews or a survey, then draft the plan, screening questions and discussion guide. Do not launch a Cookiy study, recruit participants or invent research findings; ask separately before any platform execution.

Editorial example based on the reviewed source; not an executed task or an upstream quotation.

Check the First Result

Before accepting the output, look for these in the transcript and the files:

  • The chosen method addresses the stated decision and participant group.
  • A plan is clearly distinguished from a launched study or completed research.
  • Synthesis claims are traceable to supplied evidence; real and synthetic responses are labeled separately.
  • Missing platform access or companion references are stated, and no unapproved external action is reported as complete.

OpenDesign provides a local workspace for coding agents. Check the installed version for skill availability and setup; this source review does not verify bundled skills or include external tools and services.

05 Open source, traceable

Source, License and Verification

Where this page's facts come from, and when they were last checked.

Author
Cookiy
Skill file
SKILL.md Content checksum recorded at verification; commit 21eea10
License
MIT From LICENSE.
Repository stars
1,537 GitHub snapshot taken 2026-09-14 for the whole cookiy-ai/user-research-skill repository. Not a rating or usage count for user-research-cookiy.
Verified by OpenDesign
2026-09-14 Source file, path, license text and star count were checked. Not bundled in OpenDesign main (checked 2026-08-28); a live run has not been recorded on this page.
The Skill's Own Description
“End-to-end user research assistant — qualitative and quantitative. Use this skill whenever the user mentions user research, user interviews, discussion guides, interview guides, research plans, qualitative research, quantitative research, user surveys, survey design, usability studies, participant recruitment, research synthesis, interview transcripts, research reports, running studies with AI, or explicitly mentions Cookiy AI. Also trigger when users want to talk to customers, conduct discovery research, create a study or survey, analyze interview data, run AI-moderated interviews, or collect survey responses. Covers the full lifecycle: planning studies, creating discussion guides, running AI-moderated interviews (real or synthetic) via Cookiy, designing and distributing surveys, and synthesizing results into reports.”

Front-matter description from SKILL.md. The full file is about 3,813 characters. Read the full file on GitHub.

User Research Cookiy

Run user-research-cookiy on a Real UX Research Task.

Download OpenDesign, import user-research-cookiy from its SKILL.md link, paste the prompt above, and read the plan before the result.

Terminal
npx skills add https://github.com/cookiy-ai/user-research-skill --skill user-research-cookiy