ai-ready-product-workflow-v2: Command for Claude Code

/pm-discovery

Start a product discovery sprint session. Maps opportunities, assumptions, and experiments using Teresa Torres' Continuous Discovery framework.

Usage

/pm-discovery [product or feature name]

If a product/feature name is provided via $ARGUMENTS, use it. Otherwise ask: "What opportunity or problem are we discovering for?"

---

What This Command Does

Activates the pm-product-discovery skill and guides a structured discovery session. The output is a discovery artifact stored in features/{feature-name}/pm/discovery.md.

This command is most useful:

• After /pm-strategy (to validate strategic assumptions)
• Before /po-pipeline (to ground requirements in validated user needs)
• When you have a hypothesis but need to structure what to test

---

Session Flow

Step 1 — Discovery Context

Ask the user (sequentially):

1. "What outcome are you trying to achieve? (business goal, not feature)"
2. "What do you currently know about users' pain in this area? Any existing research or data?"
3. "What are the key assumptions you're making that, if wrong, would invalidate this direction?"
4. "Have you talked to any users? If yes, what did you hear?"

Accept incomplete answers. The Opportunity Solution Tree will surface gaps.

Step 2 — Opportunity Solution Tree (Teresa Torres)

Activate pm-product-discovery skill. Read pm-framework/product-discovery/rules.md before generating output.

Build the OST in this structure:

Desired Outcome (business metric)
└── Opportunity 1: [unmet user need / pain / job]
    ├── Opportunity 1a: [more specific need]
    │   ├── Solution A: [potential solution]
    │   └── Solution B: [potential solution]
    └── Opportunity 1b: [more specific need]
        └── Solution C: [potential solution]
└── Opportunity 2: [unmet user need]
    └── ...

Rules for the OST:

• Opportunities = user needs, pains, or desires (not solutions, not features)
• Solutions = concrete product ideas that address one or more opportunities
• Start broad, then drill down to specific sub-opportunities
• Aim for 2–4 top-level opportunities, 2–3 sub-opportunities each