Examples of Products Built Using pm-product-discovery Skills: Real-World Implementation Guide

The pm-product-discovery plugin provides a /discover command and modular skills that enable product teams to run structured discovery workflows on products ranging from AI writing assistants to SaaS notification systems.

The pm-product-discovery skills in the phuryn/pm-skills repository offer a systematic approach to product validation through orchestrated commands and reusable skill modules. These tools help product managers explore ideas, surface risky assumptions, and design experiments for both new ventures and existing product enhancements.

What Are pm-product-discovery Skills?

The plugin bundles discrete capabilities—ideation, assumption mapping, prioritization, and experiment design—into a cohesive workflow. At the center is the /discover command defined in pm-product-discovery/commands/discover.md, which automatically strings together the appropriate skills based on whether you're building something new or improving an existing product.

Individual skills reside in the pm-product-discovery/skills/ directory, each with its own SKILL.md file defining prompts and expected outputs. These include brainstorm-ideas-existing, identify-assumptions-existing, prioritize-assumptions, and brainstorm-experiments-existing for iterative development, alongside corresponding "new" variants for greenfield projects.

Real-World Product Examples

The repository documentation includes concrete invocations demonstrating how the discovery workflow applies to different product contexts.

AI Writing Assistant for Non-Native Speakers

The README.md file showcases an AI writing assistant designed to help non-native speakers improve their language skills. The example prompt asks "What are the riskiest assumptions for our AI writing assistant idea?" and demonstrates the command invocation /discover New product: AI writing assistant for non-native speakers. This triggers the full discovery pipeline for a new product, generating multi-perspective ideas, identifying risky assumptions, and proposing validation experiments.

Smart Notification System for SaaS Tools

For existing product enhancements, the documentation in pm-product-discovery/commands/discover.md includes the example /discover Smart notification system for our project management tool. This represents a feature addition aimed at reducing alert fatigue for existing users. The command automatically selects the "existing" variant of skills, analyzing the current product context to brainstorm relevant notification improvements.

Churn Reduction Experiments

The repository also includes a growth-focused example targeting user retention: "reduce churn in our onboarding flow". This invocation triggers the brainstorm-experiments-existing skill to design quick-validation tests such as A/B experiments, fake door tests, or targeted user interviews specifically for the existing onboarding funnel.

The Discovery Workflow Pipeline

When you invoke the /discover command with a product idea, the system executes a six-stage pipeline defined in the command implementation:

1. Context capture – Determines whether the idea targets an existing or new product.
2. Brainstorming – Runs brainstorm-ideas-existing or brainstorm-ideas-new to generate perspectives from PM, Designer, and Engineer viewpoints.
3. Assumption identification – Calls identify-assumptions-existing or identify-assumptions-new to surface risky assumptions.
4. Assumption prioritization – Uses prioritize-assumptions to map assumptions on an Impact × Risk matrix.
5. Experiment design – Invokes brainstorm-experiments-existing or brainstorm-experiments-new to propose validation methods.
6. Discovery plan generation – Compiles outputs into a markdown-formatted plan ready for stakeholder review.

How to Run Discovery on Your Products

You can execute the complete workflow using the /discover command followed by your product description. For new products, the system automatically selects the "new" skill variants.

/discover AI-powered meeting summarizer for remote teams

For existing product enhancements, the command detects the context and uses the "existing" skill variants:

/discover Smart notification system for our project management tool

To focus specifically on experimentation for existing features:

/discover Reduce churn in our onboarding flow

Chaining Individual Skills

If you prefer granular control over the discovery process, invoke skills individually rather than using the orchestrated command. Each skill file (such as brainstorm-ideas-existing/SKILL.md and prioritize-assumptions/SKILL.md) defines specific inputs and outputs.

brainstorm-ideas-existing "Project management dashboard"
identify-assumptions-existing "Dashboard feature set"
prioritize-assumptions "Dashboard assumptions"
brainstorm-experiments-existing "Top-ranked dashboard assumptions"

This manual chaining allows you to iterate on specific stages or insert custom analysis between steps.

Summary

• The pm-product-discovery plugin in phuryn/pm-skills provides both a unified /discover command and modular skills for product discovery.
• Real-world examples include an AI writing assistant for language learners, a smart notification system for project management tools, and churn reduction experiments for onboarding flows.
• The workflow automatically selects appropriate skills based on whether you're building a new product or enhancing an existing one.
• Skills are defined in individual SKILL.md files within pm-product-discovery/skills/ directories, making them transparent and customizable.
• You can run the full automated pipeline or manually chain individual skills like brainstorm-ideas-existing and prioritize-assumptions for custom workflows.

Frequently Asked Questions

What types of products work best with pm-product-discovery skills?

The skill set handles both greenfield products and existing product enhancements. The repository demonstrates usage with AI-powered consumer apps (like writing assistants), B2B SaaS features (like notification systems), and growth experiments (like onboarding optimizations). The system automatically adjusts its approach based on whether you specify a "new product" or an existing feature context.

Can I run individual skills without using the full /discover command?

Yes. Each skill is independently invokable. For example, you can run brainstorm-ideas-existing or identify-assumptions-new separately without triggering the full six-stage pipeline. This is useful when you want to focus on a specific discovery phase or iterate on assumptions before designing experiments.

How does the assumption prioritization work in practice?

The prioritize-assumptions skill maps identified assumptions onto an Impact × Risk matrix, helping teams identify "leap-of-faith" assumptions that require immediate validation. According to the skill definition in pm-product-discovery/skills/prioritize-assumptions/SKILL.md, this process ranks assumptions by their potential to invalidate the product hypothesis if proven wrong.

Where are the skill definitions and command logic stored?

The main command logic resides in pm-product-discovery/commands/discover.md. Individual skill definitions are stored in subdirectories under pm-product-discovery/skills/, such as brainstorm-ideas-existing/SKILL.md, identify-assumptions-existing/SKILL.md, and brainstorm-experiments-existing/SKILL.md. The README.md file in the repository root provides high-level usage examples and context.

{
  "mcpServers": {
    "instagit": {
      "command": "npx",
      "args": ["-y", "instagit@latest"]
    }
  }
}