How the `/discover` Command Executes a Product Discovery Cycle in phuryn/pm-skills

The /discover command orchestrates a complete product discovery cycle by chaining modular AI skills—from context gathering and idea brainstorming to assumption prioritization and experiment design—into a single interactive workflow that generates a markdown discovery plan.

The phuryn/pm-skills repository provides a modular framework for product management workflows, where the /discover command serves as the primary interface for running structured discovery cycles. This command integrates multiple specialized skills to transform a raw prompt into a validated discovery plan, enabling product managers to move from ambiguous problem statements to testable hypotheses within a single session.

The Architecture of a Slash Command-Driven Discovery Cycle

The /discover command defined in pm-product-discovery/commands/discover.md implements a seven-stage workflow. Each stage invokes a specific skill (a markdown-driven AI assistant) that handles a discrete part of the discovery process, creating a pipeline that balances thoroughness with speed.

Context Gathering for Existing vs. New Products

The workflow begins by determining the discovery context. The command distinguishes between continuous discovery for existing products and initial discovery for new ventures. It prompts the user to clarify the discovery question, any known data, and the specific decisions the discovery will inform. This checkpoint ensures the subsequent stages target the appropriate risk profile.

Idea Brainstorming Phase

Next, the command invokes either brainstorm-ideas-existing or brainstorm-ideas-new skills (defined in pm-product-discovery/skills/brainstorm-ideas/SKILL.md). These skills generate ten cross-functional ideas spanning product management, design, and engineering perspectives. The user then selects 3–5 ideas to carry forward, filtering the solution space before deeper analysis.

Assumption Identification and Risk Categories

For each selected idea, the command deploys identify-assumptions-existing or identify-assumptions-new skills from pm-product-discovery/skills/identify-assumptions/SKILL.md. These skills surface assumptions across five risk categories:

• Value: Will customers benefit?
• Usability: Can users navigate the solution?
• Feasibility: Can we build it technically?
• Viability: Can we sustain it business-wise?
• Go-to-Market: Can we reach and acquire customers (new products only)?

Assumption Prioritization Matrix

The prioritize-assumptions skill (located at pm-product-discovery/skills/prioritize-assumptions/SKILL.md) maps every identified assumption on an Impact × Risk matrix. This stage highlights "leap-of-faith" assumptions—those with high potential impact but high uncertainty—and ranks them for immediate testing.

Experiment Design and Validation

The brainstorm-experiments-existing or brainstorm-experiments-new skills from pm-product-discovery/skills/brainstorm-experiments/SKILL.md design concrete validation experiments for each prioritized assumption. Proposed methods include A/B tests, fake doors, prototypes, pretotypes, and landing pages, each defined with specific success criteria, effort estimates, and timelines.

Discovery Plan Assembly and Next Steps

All outputs compile into a markdown discovery plan saved to the user's workspace. This document records the initial context, generated ideas, selected ideas, critical assumptions, experiment matrix, and short-term timeline. Finally, the command suggests optional follow-up actions, such as creating a PRD, drafting interview scripts, or setting up metrics.

Interactive Workflow and User Checkpoints

The discovery cycle is designed as an interactive, time-boxed process requiring approximately 15–30 minutes. Each stage functions as a checkpoint where users can redirect, skip, or dive deeper into specific areas. This architecture prevents analysis paralysis while ensuring rigorous coverage of the discovery space.

# Basic invocation for new product discovery

/discover AI writing assistant for non-native speakers

# Continuous discovery for existing products

/discover Smart notification system for our project management tool

Source Code Structure and Modularity

The workflow's modularity allows teams to customize the discovery cycle without rewriting the entire command. Swapping a skill—such as replacing the default brainstorming module with a custom variant—requires only updating the skill reference in pm-product-discovery/commands/discover.md. This design follows the repository's philosophy of composable, markdown-driven AI assistants.

Summary

• The /discover command in phuryn/pm-skills executes a seven-stage product discovery cycle by orchestrating specialized skills.
• Key stages include context gathering, idea brainstorming, assumption identification across five risk categories, Impact × Risk prioritization, and experiment design.
• Each stage serves as an interactive checkpoint, keeping the workflow time-boxed to 15–30 minutes.
• The architecture is modular—skills are defined in separate markdown files (e.g., pm-product-discovery/skills/identify-assumptions/SKILL.md) and can be swapped or updated independently.
• Output is a comprehensive markdown discovery plan with optional next-step suggestions for PRDs, interviews, or metrics.

Frequently Asked Questions

How does the /discover command handle different product maturity levels?

The command automatically branches based on product maturity. For existing products, it invokes the brainstorm-ideas-existing and identify-assumptions-existing skills to focus on continuous improvement and feature expansion. For new products, it uses the brainstorm-ideas-new and identify-assumptions-new variants, which include the additional Go-to-Market risk category critical for initial validation.

What types of experiments does the /discover command suggest?

According to the brainstorm-experiments skill definitions in pm-product-discovery/skills/brainstorm-experiments/SKILL.md, the command suggests validation methods including A/B tests, fake door tests, prototypes, pretotypes, and landing pages. Each experiment includes defined success criteria, effort estimates, and timelines appropriate for testing the specific "leap-of-faith" assumptions identified in the prioritization stage.

Can I customize the skills used in the discovery cycle?

Yes. The architecture is intentionally modular as implemented in pm-product-discovery/commands/discover.md. You can replace any skill reference—such as substituting a custom brainstorming module for the default brainstorm-ideas skill—without modifying the core command logic. This allows teams to adapt the discovery cycle to their specific industry context or methodological preferences.

How long does a typical /discover session take?

The workflow is designed to be time-boxed to approximately 15–30 minutes. Each stage includes interactive checkpoints where users can make rapid decisions (selecting 3–5 ideas from ten, prioritizing assumptions on a matrix) or choose to skip sections, ensuring the process remains lightweight while still delivering a complete discovery plan.

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