Goal: Transform customer interview transcripts into structured summaries organized by Jobs to Be Done framework, satisfaction signals, and action items.Documentation Index
Fetch the complete documentation index at: https://docs.getcore.me/llms.txt
Use this file to discover all available pages before exploring further.
Tools Required
This skill runs using CORE memory only. No integrations required.Step 1: Parse the Interview Input
Accept a transcript (pasted, uploaded, or linked). Extract metadata:- Interview date
- Participant name(s)
- Product area or context
Step 2: Extract Interview Sections
Parse the transcript into these categories:- Background — Customer context (role, company, situation)
- Current Solution — How they currently solve the problem
- What They Like — Positive aspects (rate each 1-5)
- Problems — Pain points and frustrations (list 3-5)
- Key Insights — Surprising findings or notable quotes (max 125 characters each, 3-5)
- Jobs to Be Done — What the customer is trying to accomplish (1-2 sentences)
Step 3: Identify Satisfaction Signals
For each problem or friction point mentioned, estimate:- Severity (1-5, where 5 is critical)
- Frequency (how often they encounter it)
- Current workaround (what they do instead)
Step 4: Synthesize Action Items
List follow-ups with:- Action
- Owner (You / Customer / Team)
- Due date (if mentioned)
Step 5: Present the Summary
Output in this format:Interview Summary 📅 Date & Participants
- Date: [Date]
- Interviewee: [Name, Role]
- Interviewer(s): [Your name]
- [Aspect]: [Rating] — [Quote or reason]
- [Problem]: Severity [1-5], Frequency [often/sometimes/rare] — [Workaround]
- [Surprising finding or direct quote, max 125 chars]
- [Action] — Owner: [You/Customer/Team] — Due: [Date]
Edge Cases
- No transcript provided: Ask the user to paste or upload the interview recording/notes. If they only have a voice recording, offer to transcribe (note: this requires external tooling).
- Partial interview: Proceed with what’s available. Note sections as “Not discussed.”
- Vague language: Ask clarifying questions in the output (“You mentioned ‘frustration’—can you tell me more about that?”).
- Multiple interviewees: Create one summary per person if their perspectives differ significantly; otherwise, blend insights and note where views diverged.
