Research Synthesis
Turn interview transcripts, surveys, and support feedback into themed insights and prioritized product opportunities for solo builders validating what to build.
Overview
Research Synthesis is an agent skill most often used in Idea (also Validate) that turns interviews, surveys, and feedback into themed insights and prioritized opportunities.
Install
npx skills add https://github.com/anthropics/knowledge-work-plugins --skill research-synthesisWhat is this skill?
- Structured output: executive summary, themed evidence with participant quotes, and insights-to-opportunities table
- Accepts interview notes, CSV surveys, usability tests, support tickets, NPS/CSAT, and app store reviews
- Pairs with the sibling user-research skill for methods; this skill focuses on synthesis and recommendations
- Argument-driven /research-synthesis invocation with study metadata (method, participants, date range)
- Output sections include Executive Summary, Key Themes (with prevalence), and Insights → Opportunities table
Adoption & trust: 1.9k installs on skills.sh; 19.6k GitHub stars; 3/3 security scanners passed (skills.sh audits).
What problem does it solve?
You collected user interviews and survey rows but lack a consistent way to find themes, evidence, and what to build next.
Who is it for?
Solo builders with raw research artifacts who need a fast, quotable synthesis before prototyping or rewriting messaging.
Skip if: Teams that only need statistical survey analysis without qualitative theming, or when you have no participant data to synthesize yet.
When should I use this skill?
You have interview transcripts, survey results, usability test notes, support tickets, or NPS responses that need distillation into patterns and prioritized next steps.
What do I get? / Deliverables
You get a structured synthesis with themes, quotes, and an insights-to-opportunities table ready to inform scope and roadmap decisions.
- Research Synthesis markdown with themes, quotes, and opportunities table
- Executive summary and implication notes per theme
Recommended Skills
Journey fit
Spans multiple journey phases - primary shelf plus alternate fits below.
User research synthesis is the canonical outcome of discovery work before you commit to a scope—where raw qualitative data becomes decisions. The skill accepts transcripts, surveys, usability notes, and reviews—the core artifacts of idea-phase research, not post-launch analytics dashboards.
Where it fits
Synthesize five discovery interviews into themes before choosing a wedge feature.
Cluster survey responses to define segments for a landing page headline test.
Map usability test pain points to a cut list for an MVP scope doc.
Mine recurring support ticket themes to prioritize fix vs FAQ vs product work.
How it compares
Use for post-collection synthesis—not as a substitute for running interviews (see user-research methods skills first).
Common Questions / FAQ
Who is research-synthesis for?
Solo and indie builders, PMs, and founders who already have research data and need themes, segments, and prioritized next steps without hiring a full UX research team.
When should I use research-synthesis?
In Idea research after interviews or surveys; in Validate scope when reconciling feedback before cutting features; anytime support tickets, NPS, or app reviews need pattern extraction before you commit to build.
Is research-synthesis safe to install?
Treat pasted transcripts and CSVs as sensitive user data; review the Security Audits panel on this Prism page before piping production customer data through an agent.
SKILL.md
READMESKILL.md - Research Synthesis
# /research-synthesis > If you see unfamiliar placeholders or need to check which tools are connected, see [CONNECTORS.md](../../CONNECTORS.md). Synthesize user research data into actionable insights. See the **user-research** skill for research methods, interview guides, and analysis frameworks. ## Usage ``` /research-synthesis $ARGUMENTS ``` ## What I Accept - Interview transcripts or notes - Survey results (CSV, pasted data) - Usability test recordings or notes - Support tickets or feedback - NPS/CSAT responses - App store reviews ## Output ```markdown ## Research Synthesis: [Study Name] **Method:** [Interviews / Survey / Usability Test] | **Participants:** [X] **Date:** [Date range] | **Researcher:** [Name] ### Executive Summary [3-4 sentence overview of key findings] ### Key Themes #### Theme 1: [Name] **Prevalence:** [X of Y participants] **Summary:** [What this theme is about] **Supporting Evidence:** - "[Quote]" — P[X] - "[Quote]" — P[X] **Implication:** [What this means for the product] #### Theme 2: [Name] [Same format] ### Insights → Opportunities | Insight | Opportunity | Impact | Effort | |---------|-------------|--------|--------| | [What we learned] | [What we could do] | High/Med/Low | High/Med/Low | ### User Segments Identified | Segment | Characteristics | Needs | Size | |---------|----------------|-------|------| | [Name] | [Description] | [Key needs] | [Rough %] | ### Recommendations 1. **[High priority]** — [Why, based on which findings] 2. **[Medium priority]** — [Why] 3. **[Lower priority]** — [Why] ### Questions for Further Research - [What we still don't know] ### Methodology Notes [How the research was conducted, any limitations or biases to note] ``` ## If Connectors Available If **~~user feedback** is connected: - Pull support tickets, feature requests, and NPS responses to supplement research data - Cross-reference themes with real user complaints and requests If **~~product analytics** is connected: - Validate qualitative findings with usage data and behavioral metrics - Quantify the impact of identified pain points If **~~knowledge base** is connected: - Search for prior research studies and findings to compare against - Publish the synthesis to your research repository ## Tips 1. **Include raw quotes** — Direct participant quotes make insights credible and memorable. 2. **Separate observations from interpretations** — "5 of 8 users clicked the wrong button" is an observation. "The button placement is confusing" is an interpretation. 3. **Quantify where possible** — "Most users" is vague. "7 of 10 users" is specific.