discovery-debrieflisted
Install: claude install-skill Layneformalized225/ai-cofounder
# Discovery Debrief
Structured extraction after a customer/prospect conversation. Every conversation is data. This skill turns conversations into actionable learnings.
**Triggers:** "talked to a customer", "customer call", "debrief", "discovery call", "had a call with...", "met with..."
---
## Step 0: Context
1. Read `MEMORY.md` — current wedge, ICP, constraint
2. Read `memory/hypotheses.json` — active hypotheses
3. Ask: "Who did you talk to? Tell me in free form."
---
## Step 1: Structured Extraction
After the founder's free-form story — extract structured data. Ask clarifying questions ONE AT A TIME (don't batch).
### 1.1 Who
- Name, role, company, team/company size
- How the contact was established (inbound/outbound/referral)
- Buyer or user? (often different people)
### 1.2 Demand Reality (is demand real?)
> "Would this person be upset if our product disappeared tomorrow?"
Look for behavioral evidence, not words:
- Paying or willing to pay? How much?
- Using the product? How often?
- Building workflow around the product?
- Would scramble if the product vanished?
**Red flags:** "interesting", "need to think about it", "show it to my colleagues" — this is politeness, not demand.
### 1.3 Status Quo (what are they doing now?)
> "How do they solve this problem today — even poorly?"
Look for the specific workflow:
- What tools/processes do they use?
- How much time/money do they spend?
- Who does it manually?
- What breaks in the current process?
**Red flag:** "t