discover — Claude Code command: what it does, how to install
Runs a structured product discovery cycle from initial ideation to experiment design. Useful for testing product hypotheses before writing code.
MIT ★ 28,211 updated 2026-10-16
/plugin marketplace add phuryn/pm-skills
/plugin install pm-product-discovery@pm-skillsWhat discover does
The command chains multiple product skills into an interactive 15-to-30-minute discovery workflow. It starts by clarifying whether you work on a brand-new concept or an existing product with active users. You can feed it context through uploaded research files, transcripts, PRDs, or simple links.
Claude Code then generates ten product concepts across PM, engineering, and design angles, maps assumptions across risk categories, and plots them on an Impact and Risk matrix. For the highest-risk assumptions, it designs validation tests such as landing pages, prototypes, or A/B tests with concrete success metrics and timelines.
Everything ends up in a structured Markdown discovery plan saved directly to your workspace. The workflow requires user decisions at several checkpoints to select ideas and assumptions, so running it fully unattended will pause execution.
What it helps with
- Brainstorm product ideas across PM, engineering, and design angles
- Surface critical assumptions across value, usability, feasibility, and viability
- Prioritize riskiest assumptions on an impact and uncertainty matrix
- Design targeted experiments with timelines, effort estimates, and success criteria
- Compile discovery findings into a structured Markdown document
Example prompts for the agent
- /discover Smart notification system for our project management tool
- /discover New product: AI writing assistant for non-native speakers
- /discover
How to install discover
1Plugin marketplace
/plugin marketplace add phuryn/pm-skills/plugin install pm-product-discovery@pm-skills2Manually: put the files in place
.claude/commands/discover.mdCall it with /discover. For all projects use ~/.claude/commands/
Manually: put the files in place
~/.codex/prompts/discover.mdCall it with /prompts:discover. Codex marks this format deprecated in favour of skills
Manually: put the files in place
.github/prompts/discover.prompt.mdManually: put the files in place
.gemini/commands/discover.tomlGemini stores commands as TOML: the prompt text goes into the prompt field
Questions about discover
How to install discover in Claude Code?
Run /plugin marketplace add phuryn/pm-skills to register the marketplace, then install the package with /plugin install pm-product-discovery@pm-skills.
What does the discover command do?
It guides you through an end-to-end product discovery workflow, moving from brainstorming ideas to ranking critical assumptions and designing validation experiments.
What is the difference between existing and new product discovery?
For new concepts, the command emphasizes desirability and tests like landing pages, whereas existing products focus on user data, A/B tests, and prototypes.
Source text
discover.md · MIT
--- description: Run a full product discovery cycle — from ideation through assumption mapping to experiment design argument-hint: "<product or feature idea>" --- # /discover -- Full Discovery Cycle Run a structured product discovery process that moves from divergent thinking to focused validation. This command chains multiple skills into a single end-to-end workflow. ## Invocation ``` /discover Smart notification system for our project management tool /discover New product: AI writing assistant for non-native speakers /discover # asks what you're discovering ``` ## Workflow ### Step 1: Understand the Discovery Context Determine whether this is: - **Existing product** — continuous discovery on a known product with real users - **New product** — initial discovery for a concept without validated demand Ask the user: - What are you exploring? (product idea, feature area, opportunity space) - What do you already know? (prior research, customer feedback, data) - What decisions will this discovery inform? (build/kill, prioritize, pivot) Accept context from uploaded files (research, PRDs, transcripts, data), links, or conversation. ### Step 2: Brainstorm Ideas (Divergent Phase) Apply the **brainstorm-ideas-existing** or **brainstorm-ideas-new** skill: - Generate ideas from PM, Designer, and Engineer perspectives - Present the top 10 ideas with brief rationale - Ask the user to select 3-5 ideas to carry forward, or accept all **Checkpoint**: "Here are 10 ideas. Which ones should we stress-test? Pick 3-5, or I can carry all forward." ### Step 3: Identify Assumptions (Critical Thinking Phase) For each selected idea, apply the **identify-assumptions-existing** or **identify-assumptions-new** skill: - Surface assumptions across risk categories: - **Value**: Will users want this? - **Usability**: Can users figure it out? - **Feasibility**: Can we build it? - **Viability**: Does the business case work? - **Go-to-Market** (new products only): Can we reach and convert users? - Use devil's advocate multi-perspective analysis - Compile a master list of all assumptions across all ideas ### Step 4: Prioritize Assumptions (Focus Phase) Apply the **prioritize-assumptions** skill: - Map assumptions on an Impact × Risk matrix - Identify the "leap of faith" assumptions — high impact, high uncertainty - Rank assumptions by test priority - Group related assumptions that can be tested together **Checkpoint**: "Here are your riskiest assumptions. Which ones feel most critical to validate first?" ### Step 5: Design Experiments (Validation Phase) For the top-priority assumptions, apply **brainstorm-experiments-existing** or **brainstorm-experiments-new** skill: - Design 1-2 experiments per critical assumption - For existing products: A/B tests, fake doors, prototypes, user tests, data analysis - For new products: XYZ hypotheses, pretotypes, landing pages, concierge MVPs - Include success criteria, timeline, and effort for each - Sequence experiments by dependency and effort ### Step 6: Create Discovery Plan Compile everything into a discovery plan document: ``` ## Discovery Plan: [Topic] **Date**: [today] **Product Stage**: [existing/new] **Discovery Question**: [what we're trying to learn] ### Ideas Explored [Summary of brainstormed ideas with brief descriptions] ### Selected Ideas for Validation [3-5 ideas carried forward with rationale] ### Critical Assumptions | # | Assumption | Category | Impact | Uncertainty | Priority | |---|-----------|----------|--------|-------------|----------| ### Validation Experiments | # | Tests Assumption | Method | Success Criteria | Effort | Timeline | |---|-----------------|--------|-----------------|--------|----------| ### Experiment Details [For each experiment: hypothesis, setup, measurement, decision criteria] ### Discovery Timeline Week 1: [experiments] Week 2: [experiments] Week 3: [analysis and decision] ### Decision Framework - If [experiment] succeeds → proceed to [next step] - If [experiment] fails → [pivot/kill/investigate further] ``` Save the plan as a markdown file to the user's workspace. ### Step 7: Offer Next Steps - "Want me to **create a PRD** for the top idea?" - "Should I **design an interview script** to supplement these experiments?" - "Want me to **set up metrics** to track the experiments?" - "Should I **estimate effort** and create user stories for the MVP?" ## Notes - This is a 15-30 minute structured workflow — let the user know upfront - At each checkpoint, the user can redirect, skip, or go deeper - If the user has research data, pull insights from it before brainstorming - The discovery plan should be a living document — offer to update it as experiments run - For new products, emphasize desirability validation before feasibility - For existing products, check if there's usage data that can inform assumptions
© Paweł Huryn · MIT · phuryn/pm-skills