ask_humans” and “the agent knows to ask three clarifying questions, then run ask_humans in claim mode with a sensible default N, then poll get_results until the study completes.”
When skills help
- Repeatable workflows. Validating positioning copy, screening Panel respondents, generating an insights report — anything you do more than twice.
- Cross-team consistency. Two PMs on different machines run the same workflow the same way because the skill ships the playbook.
- Onboarding. A new teammate installs the skills and immediately runs the team’s standard research patterns.
How skills work
A skill is a Markdown file with YAML frontmatter. The frontmatter tells the agent when the skill applies; the body tells it how to execute. Example shape:Two ways to install
- Claude Code
- Cursor
- Claude Desktop
Drop the skill file into your project’s Claude Code auto-discovers skills on next session start. Verify with
.claude/skills/ directory (or ~/.claude/skills/ for global use):/skills.What’s in the library
The skills library ships ready-to-use skills for the most common User Intuition workflows:validate-claim-with-humans
Test whether a marketing claim is believable to real respondents.
run-preference-study
Compare 2-5 options (taglines, designs, value props) head-to-head.
recruit-with-screeners
Build screener questions to recruit a specific Panel segment.
analyze-fielded-study
Pull results, summarize themes, surface verbatim quotes.
watch-running-studies
Daily digest of fielding studies — runs in cron via the CLI.
Skill vs MCP vs CLI — which goes where
Skills sit on top of MCP, MCP and CLI sit on top of the same tool surface, and everything sits on top of the REST API. Pick the highest level that gets the job done.
Next steps
Skills library
Browse and copy ready-to-use skills.
MCP quickstart
Skills require an MCP-connected client. Wire one up first.

