name: discover
description: "Run a full discovery cycle from opportunity mapping to experiment design. Routes to the continuous-discovery skill."
argument-hint: "[describe your product, team goal or outcome, and any existing customer insights or research]"

Discover

This skill routes to continuous-discovery, which provides a complete continuous discovery framework based on Teresa Torres' Opportunity Solution Trees and Marty Cagan's dual-track agile.

When to use this:
- Starting a discovery sprint for a new outcome or problem space
- Moving from a vague "we should improve X" to a specific, testable opportunity
- Mapping the full opportunity space before committing to a solution
- Answering "how would you run discovery for this product area?" in PM interviews

What continuous-discovery covers:
- Opportunity Solution Tree (OST): outcome → opportunities → solutions → experiments
- Desired outcome definition: framing the outcome as a measurable business result, not a feature goal
- Opportunity mapping: extracting opportunities from customer interviews, behavioral data, and support tickets
- Opportunity clustering: grouping raw insights into themes and sub-themes
- Opportunity scoring: intensity × frequency × strategic fit × confidence
- OST pruning: selecting the top opportunity to focus on and why
- Solution generation: 5+ diverse candidates before converging (avoiding anchoring)
- Assumption mapping: what must be true for this solution to work? (desirability, viability, feasibility)
- Experiment design: pretotypes, concierge MVP, Wizard of Oz, landing page tests
- Dual-track agile (Marty Cagan): discovery and delivery running in parallel, not sequentially
- Discovery cadence: weekly customer touchpoints, bi-weekly OST updates, monthly outcome review
- Common discovery mistakes: skipping opportunity mapping, jumping to the first solution, testing in isolation

The skill will:
1. Ask for the desired outcome and any existing customer insights
2. Define the outcome in measurable terms
3. Build an opportunity map from available insights (or help design research to gather them)
4. Score and select the top opportunity to focus on
5. Generate 5+ solution candidates for that opportunity
6. Map assumptions for the top solution candidates
7. Design one pretotype experiment for the riskiest assumption

After completing, suggest: "Use /interview to prepare your next customer interview, or /prioritize to score and rank the full opportunity list."

Load the full skill: continuous-discovery

Apply your product and discovery context to: $ARGUMENTS