Create a PRD
The scenario
You need to write a PRD for a feature. Normally this means 4-6 hours of staring at a blank doc, trying to remember what sections to include, formatting tables, and anticipating what the CTO will ask. With /prd-generator, you answer Claude's questions about the problem and get a complete first draft in 15 minutes.
Example output
PRD: Mobile Checkout Redesign
Status: Draft · Owner: Ron · Date: Jan 30, 2026
Problem Statement Enterprise users abandon checkout at 3x the rate of SMB users. The current 5-step checkout flow requires information that enterprise buyers don't have on hand — billing codes, PO numbers, and department cost centers. This causes 67% cart abandonment for enterprise accounts vs. 23% for SMB.
Evidence
- Cart abandonment: 67% enterprise vs. 23% SMB (Mixpanel, last 90 days)
- Support tickets: 34 tickets/month mentioning "checkout" or "billing confusion"
- User interviews: 4 of 6 enterprise users cited "too many steps" unprompted
- Revenue impact: Est. $180k/quarter in lost enterprise upgrades
Proposed Solution Reduce enterprise checkout from 5 steps to 2:
- Auto-populate billing details from company profile
- "Complete Later" option for PO numbers — don't block the purchase
- Saved payment methods at the organization level
Success Metrics
- Reduce enterprise cart abandonment from 67% to below 40%
- Reduce checkout-related support tickets by 50%
- No increase in payment failures or chargebacks
Risks and Mitigations
Risk Likelihood Impact Mitigation Billing errors from auto-populate Medium High Add confirmation step before final charge Security concerns with saved payments Low High SOC2 compliance review before launch PO numbers never get added ("Complete Later" becomes "Complete Never") Medium Medium Auto-reminder emails at 7, 14, 21 days Timeline
- Week 1-2: Design + prototype
- Week 3-4: Engineering build
- Week 5: QA + staged rollout (10% → 50% → 100%)
Out of Scope
- SMB checkout changes (not broken)
- Payment provider migration (separate initiative)
- International billing formats (Phase 2)
How to create this in 15 minutes
Step 1. Make sure your Product Management OS folder (second-pm-os) is open in Claude.
Step 2. In the chat panel, type:
/prd-generator mobile checkout redesign
Replace "mobile checkout redesign" with a brief description of your feature.
Step 3. Claude asks you a series of problem-first questions. This is intentional — good PRDs start with the problem, not the solution. You'll be asked things like:
- What problem are you trying to solve?
- Who experiences this problem?
- How do you know it's a real problem? (What evidence do you have?)
- Why is it important to solve now?
- What does success look like?
Answer each question. Rough notes are fine — you don't need polished prose. Claude handles the formatting.
Step 4. Claude generates a complete PRD using your context files + your answers and saves it to:
specs/outputs/[feature-name]-prd.md
Step 5. Review the draft. The first version is a starting point, not a final document.
If something doesn't look right
Tell Claude what to change. For example:
- "The risks section is too vague — add specific technical risks for the auto-populate feature"
- "Add a section about dependencies — we need the design system update to land first"
- "The timeline is too aggressive — double the engineering estimate"
- "Rewrite the problem statement for a non-technical audience"
- "Add competitive context — Stripe's checkout does this in 1 step already"
Pro tip
After your PRD looks good, run /stakeholder-simulator to simulate feedback from your CTO, UX lead, sales team, and exec. Fix the obvious concerns before your real meeting. See Get Stakeholder Feedback for the walkthrough.
What's next?
- Get Stakeholder Feedback — Simulate CTO, UX, and exec feedback before your real review
- Analyze Multiple Competitors at Once — Build a competitive landscape with comparison matrix