How do I install the Research Synthesis skill?

Run npx skillmds@latest add dragoon0x/research-synthesis in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is Research Synthesis safe to use?

This skill has not completed SkillMD's automated safety review yet. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Research Synthesis?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Research Synthesis free to use?

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

Who published Research Synthesis?

dragoon0x (@dragoon0x) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Research Synthesis by dragoon0x — Agent Skill

Name: Research Synthesis
Author: dragoon0x

Research Synthesis

Synthesize user interviews, surveys, support tickets, or other research into themes and recommendations. Use whenever the user shares interview transcripts, survey responses, support data, customer research, or asks for synthesis of qualitative input.

dragoon0x 1b54621 2.4 KB Updated Sep 17, 2026

File contents

research-synthesis

raw research is data. synthesis is making it useful for decisions.

the process

1. read everything before synthesizing

never synthesize from a sample if the full set is available. patterns you notice in the first three interviews are often artifacts of the order you read them in.

2. tag, don't paraphrase

go through and tag statements. tags include:

job-to-be-done (what they're trying to accomplish)
frustration (what's blocking them)
workaround (what they do today)
desire (what they wish existed)
hard constraint (what they can't change)
non-goal (something explicitly not important to them)

keep the original wording. tag it. don't rewrite it yet.

3. cluster

group tagged statements by the shape of the underlying need, not the surface vocabulary. "i hate spreadsheets" and "i wish there was a faster way to track this" might be the same cluster.

4. count

for each cluster:

how many distinct participants voiced this
how strongly (passing mention vs. animated frustration)
whether it came up unprompted or only when probed

5. theme

a theme is a cluster with a name. name it in the user's language, not your team's.

output shape

theme: [short label]
size: [N participants]
strength: [low / medium / high]
representative quotes: [2-3, attributed to participant id]
underlying job: [what they're trying to accomplish]
current workaround: [what they do today]
opportunity: [what a product could do here]

→ implication: [what this means for product decisions]
→ confidence: [low / medium / high based on sample and consistency]

what to avoid

inventing quotes
generalizing from one passionate participant
mixing themes (one theme per cluster)
jumping to solutions before themes are stable
using softening adverbs (most, many, some) without numbers behind them
presenting themes in a deck that has the conclusion before the evidence

how to know synthesis is done

a good synthesis lets a reader who wasn't in the interviews make the same product decision you would. if they'd reach a different conclusion, the synthesis isn't complete or the evidence isn't strong enough.

---name: research-synthesisdescription: Synthesize user interviews, surveys, support tickets, or other research into themes and recommendations. Use whenever the user shares interview transcripts, survey responses, support data, customer research, or asks for synthesis of qualitative input.---# research-synthesisraw research is data. synthesis is making it useful for decisions.## the process### 1. read everything before synthesizingnever synthesize from a sample if the full set is available. patterns you notice in the first three interviews are often artifacts of the order you read them in.### 2. tag, don't paraphrasego through and tag statements. tags include:- job-to-be-done (what they're trying to accomplish)- frustration (what's blocking them)- workaround (what they do today)- desire (what they wish existed)- hard constraint (what they can't change)- non-goal (something explicitly not important to them)keep the original wording. tag it. don't rewrite it yet.### 3. clustergroup tagged statements by the shape of the underlying need, not the surface vocabulary. "i hate spreadsheets" and "i wish there was a faster way to track this" might be the same cluster.### 4. countfor each cluster:- how many distinct participants voiced this- how strongly (passing mention vs. animated frustration)- whether it came up unprompted or only when probed### 5. themea theme is a cluster with a name. name it in the user's language, not your team's.## output shape```theme: [short label]size: [N participants]strength: [low / medium / high]representative quotes: [2-3, attributed to participant id]underlying job: [what they're trying to accomplish]current workaround: [what they do today]opportunity: [what a product could do here]→ implication: [what this means for product decisions]→ confidence: [low / medium / high based on sample and consistency]```## what to avoid- inventing quotes- generalizing from one passionate participant- mixing themes (one theme per cluster)- jumping to solutions before themes are stable- using softening adverbs (most, many, some) without numbers behind them- presenting themes in a deck that has the conclusion before the evidence## how to know synthesis is donea good synthesis lets a reader who wasn't in the interviews make the same product decision you would. if they'd reach a different conclusion, the synthesis isn't complete or the evidence isn't strong enough.

Frequently asked questions