Works with the AI Agents you already use
Claude Code
Cursor
Codex CLI
Muse
Dots+21 moreClaude CodeCursorCodex CLIMuseDots
Manus
Grok
Grok Bot+18 agentsCustomer Interviews → Themes, Jobs + Next Questions
Unlike free "summarise my interviews" prompts: evidence-coded themes with participant IDs, contradictions kept, opportunity hypotheses labelled as hypotheses — no invented quotes or market proof.
$7
Customer Interviews → Themes, Jobs + Next Questions
Example session with this skill installed
Analyze these 3 transcripts. Research question: Why do designers stop using our component library after two weeks? Participants: P01 (Senior), P02 (Junior), P03 (Freelance). Lens: Friction, Trust.
- Read your context and instructions
- Compiled the customer interviews themes
I have synthesized the three transcripts provided. The analysis identifies a significant tension between initial setup speed and long-term maintenance trust.
Customer interview synthesis: Designer retention for component library
Research question: Why do designers stop using our component library after two weeks?
Study scope: 3 qualitative interviews with varied seniority levels.
Decision context: Identifying barriers to long-term adoption.
Participants: P01, P02, P03
Generated: 2024-05-22 · Evidence synthesis for human verification — not market proof
1. Executive readout
What the interviews suggest: The library is easy to start with but lacks the flexibility needed for edge cases, leading designers to "break" components and lose sync.
- What they do not establish: Whether this leads to total churn or just selective use of specific atoms.
Most important implication: The "all or nothing" nature of the library causes trust decay when a single component fails a specific use case.
- Evidence confidence: Medium; consistent across 3 varied roles but needs a larger sample to confirm prevalence.
4. Evidence table
| Evidence ID | Participant | Location | Observation / paraphrase | Direct quote | Code(s) |
|---|---|---|---|---|---|
| E1 | P01 | p.2 | Overriding props feels "hacky" and brittle. | “I felt like I was fighting the library instead of using it.” | Friction |
| E2 | P02 | p.4 | Documentation is clear for basics but missing complex states. | “I couldn't find the hover state for the nested dropdown.” | Friction |
| E3 | P03 | p.1 | Fear of library updates breaking custom workarounds. | “I’m scared to pull the latest version because I’ve detached so much.” | Trust |
5. Themes and patterns
Theme 1: The "Detachment Trap"
Pattern: Users start by using components as intended but detach them as soon as a non-standard requirement appears.
- Evidence: E1, E3
- Variation: P02 (Junior) detaches less frequently but expresses more frustration when stuck.
- Interpretation: High initial velocity is offset by a "wall" encountered at the 14-day mark of a project.
8. Opportunity hypotheses
| Priority | Hypothesis | Evidence basis | Smallest validation step | Signal to watch |
|---|---|---|---|---|
| 1 | If we provide a "slot" system for custom sub-components, then designers will detach 50% less often because they won't feel they are "fighting" the library. | E1, E3 | Beta test a "flexible card" component with 5 designers. | Detachment rate in Figma analytics. |
Next steps
- Review the "Detachment Trap" findings with the design systems engineering team.
- Audit the current library for components with the highest detachment rates.
- Schedule follow-up interviews focusing specifically on the handoff between Figma and React.
Connects securely to your tools. The creator never sees your data.
About this skill
Unlike free "summarise my interviews" prompts, this skill synthesises YOUR pasted transcripts or notes into an evidence-led readout: coding table, themes tied to participant IDs, jobs/pains/workarounds, contradictions, and opportunity hypotheses labelled as hypotheses. Quotes stay verbatim; sample size and confidence stay visible. No invented customers, prevalence, revenue or ROI claims.
Sample output (excerpt, 6 agency handoff interviews):
| Theme | Pattern | Evidence | Confidence |
|---|---|---|---|
| Handoff docs feel like busywork | Leads skip templates when the client is waiting | I2, I4, I5 (P02, P04) | medium |
| Trust breaks at first revision | Owners re-write the brief themselves after one bad handoff | I1, I6 | high |
| Job / moment | Pain | Workaround | Evidence |
|---|---|---|---|
| Start a new client project | Template fields do not match how the team actually works | Paste last project's Notion page and edit | I3, I5 |
| Client asks "what's next?" | No shared checklist the PM and designer both trust | Slack thread + screenshot dump | I2, I4 |
Opportunity hypothesis (not a fact): If the first handoff is a 5-field checklist filled in the kickoff call, then missed steps drop — validate with 3 more interviews + one observed kickoff.
Works in Claude Code, Cursor and Codex · paste transcripts or notes with participant IDs · evidence-only output · v1.0, updated Oct 2026 · questions answered via Agensi messages · More Cygnix Labs tools: https://apify.com/cygnix_labs
Stop making product decisions based on gut feelings or hallucinated customer needs. Raw interview transcripts are often messy, disorganized, and full of noise, making it difficult to extract actionable evidence without bias. This skill provides a rigorous framework for transforming raw qualitative data into structured research synthesis.
What it does
- Evidence extraction identifies specific behaviors, triggers, and workarounds directly from transcripts.
- Systematic coding applies a consistent lens to categorize data into jobs, pains, and objections.
- Pattern matching clusters evidence into themes while explicitly highlighting contradictions and negative cases.
- Opportunity mapping generates testable hypotheses linked to specific participant IDs and quotes.
- Integrity checking ensures every claim is backed by supplied text without inventing market-wide outcomes.
How it works
- Define the frame by providing your research question, study scope, and raw interview text or notes.
- Review the inventory as the agent maps the corpus, identifying completeness and sample caveats.
- Analyze the synthesis through structured tables covering coding, evidence, and tension points.
- Action the hypotheses by using the generated validation steps and next research questions to guide your sprint.
Frameworks & tools
This skill is optimized for qualitative research methodologies including Jobs-to-be-Done (JTBD), thematic analysis, and evidence-led product discovery.
Why this beats prompting it yourself
Standard prompts often produce generic summaries that lose the nuance of participant quotes or ignore evidence that contradicts the main narrative. This skill enforces a strict separation between direct evidence and interpretation, preventing the AI from claiming product-market fit or inventing ROI metrics from a small sample.
Use cases
- Product discovery to identify why users are abandoning specific workflows or features.
- Messaging optimization by extracting verbatim triggers and objections from prospect calls.
- Service design to map friction points and current workarounds in existing user journeys.
- Churn analysis to synthesize exit interviews into prioritized opportunity hypotheses.
Known limitations
Does not support audio files directly; requires text transcripts or notes. Accuracy depends on the quality and honesty of the supplied source material.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before any payment.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Money-back guarantee
Doesn't do what its page promises? You get your money back.
Trust & safety
Security scan
Security scan passed 1 month ago