Agent skill

Research Synthesis Skill: Themes and Insights From Data

The research synthesis skill turns interview transcripts, survey results, support tickets and usability notes into themes, quotes, user segments and prioritized recommendations.

Install
claude plugin marketplace add anthropics/knowledge-work-plugins
View on GitHub

Stars, license and last push from GitHub. Install count as reported by the open skills directory.

What is the research synthesis skill?

Research Synthesis is a skill in the design plugin of anthropics/knowledge-work-plugins, run as /research-synthesis. It accepts transcripts, surveys, usability notes, tickets, NPS and app store reviews, and returns a fixed report: executive summary, themes with evidence, insight to opportunity table, segments and ranked recommendations.

It is a slash command skill in the design plugin. You run /research-synthesis with your data, and the agent distills it into patterns, user segments and next steps in one set report format.

Planning a study is handled by a sibling skill, user-research. This one starts after the data exists and its job is to turn raw material into findings someone can act on.

What goes in and what comes out

  • Input: interview transcripts or notes, survey results, usability test notes, support tickets, NPS or CSAT responses and app store reviews.
  • Report header with method, participant count, date range and researcher.
  • Key themes, each with prevalence such as X of Y participants, a summary, quotes and an implication.
  • An insights to opportunities table rated by impact and effort.
  • User segments, ranked recommendations and open questions for further research.

Who the research synthesis skill is for

Best for

  • A pile of interview notes that needs themes by end of day.
  • Combining survey comments with support tickets into one view.
  • Giving a team a consistent research readout format.

Not for

  • Collecting data. The user research skill in the same plugin covers planning and interview guides.
  • Statistical proof. The output labels prevalence, but judging sample quality is still your call.
  • Visual design review. For a screen, use the Design Critique skill.

Tips the skill gives for better syntheses

  • Include raw quotes, because direct words make an insight credible.
  • Separate what you saw from what you infer: five of eight users clicked the wrong button is a fact, the layout confuses people is a reading.
  • Quantify. Seven of ten users beats most users.