How to Use the pm-product-discovery Plugin: A Complete Guide for Product Managers

The pm-product-discovery plugin is a Claude-compatible extension that exposes slash commands for structured product discovery workflows, allowing you to execute skills like interview summarization and feature prioritization by typing commands such as /pm-product-discovery:interview in your chat interface.

The pm-product-discovery plugin is part of the open-source phuryn/pm-skills repository and provides product managers with a library of markdown-based skills for discovery tasks. Unlike traditional code plugins, this extension uses declarative markdown files stored in the repository to define capabilities, making it lightweight, version-controllable, and easy to customize without redeploying code.

Plugin Architecture and File Structure

Understanding the file layout helps you locate specific capabilities and extend the plugin. The repository organizes components into four distinct layers:

• Plugin manifest – Located at pm-product-discovery/.claude-plugin/plugin.json, this file registers the plugin with Claude and supplies metadata including name, version, description, and available commands.
• Skills – Self-contained markdown definitions stored in pm-product-discovery/skills/ (e.g., summarize-interview/SKILL.md, prioritize-features/SKILL.md). Each skill contains a YAML header, human-readable description, step-by-step instructions, and optional "Further Reading" links.
• Commands – Mapping files in pm-product-discovery/commands/ (e.g., interview.md, brainstorm.md) that associate slash commands with specific skills and provide UI hints.
• README – The pm-product-discovery/README.md serves as a catalog of all available skills and commands for quick reference.

When Claude loads the plugin, it reads the manifest to expose the listed commands. Each command triggers the corresponding skill, which executes a guided process involving data collection, framework application (such as ICE or RICE scoring), and structured markdown output generation.

Available Commands and Practical Usage

The plugin exposes several slash commands under the pm-product-discovery namespace. Below are the primary commands with implementation details from the source code.

Summarize Customer Interviews

To generate a structured summary from a customer interview transcript, type:

/pm-product-discovery:interview

Claude prompts you to upload the transcript (plain text, PDF, or pasted excerpt). The command invokes the summarize-interview skill defined in summarize-interview/SKILL.md, which processes the input through a template producing a markdown document containing Date, Participants, Current Solution, Problems, Key Insights, and Action Items.

Prioritize Feature Backlogs

For backlog prioritization using Impact/Effort frameworks:

/pm-product-discovery:brainstorm

Specify that you want to prioritize features, and Claude switches to the prioritize-features skill (see prioritize-features/SKILL.md). Provide your feature list (optionally attaching a spreadsheet), and the skill evaluates each item on Impact, Effort, Risk, and Strategic Alignment, returning a ranked table of the top five recommendations.

Run a Full Discovery Cycle

Orchestrate a complete discovery workflow with:

/pm-product-discovery:discover

This high-level command chains multiple skills sequentially:

1. brainstorm-ideas-new – Generates product ideas from PM, Designer, and Engineer perspectives.
2. identify-assumptions-new – Surfaces risky assumptions underlying the ideas.
3. prioritize-assumptions – Ranks assumptions by Impact × Risk.
4. brainstorm-experiments-new – Designs experiments for the top assumptions.
5. metrics-dashboard – Defines a metrics dashboard for the downstream product.

The output comprises multiple markdown artefacts: an idea list, assumption matrix, experiment plan, and dashboard design.

Create Interview Scripts

When you need a structured interview guide rather than a summary:

/pm-product-discovery:interview

Tell Claude you need an interview script, and the plugin invokes the interview-script skill (found in interview-script/SKILL.md). You receive a structured script with warm-up sections, JTBD (Jobs-to-be-Done) probing questions, and wrap-up protocols ready to copy into a document or meeting invite.

Set Up Metrics Dashboards

To define tracking infrastructure:

/pm-product-discovery:setup-metrics

Claude requests your product's North Star metric and supporting health metrics. The metrics-dashboard skill (see metrics-dashboard/SKILL.md) returns a dashboard layout, data source suggestions, and alert thresholds in markdown format.

How the Plugin Works Step-by-Step

The execution flow follows a consistent pattern across all skills:

1. Discovery – Claude displays available slash commands under the pm-product-discovery namespace in the chat UI.
2. Invocation – You select a command (e.g., /pm-product-discovery:discover).
3. Data collection – The skill prompts for required artefacts (transcripts, feature lists, product briefs).
4. Processing – The skill applies its prescribed framework (Opportunity Score, ICE scoring, etc.) and generates a structured markdown output.
5. Result delivery – Claude returns the markdown document, which you can save, copy, or embed directly into a product backlog or wiki.

Because all skills are stored as markdown files in the phuryn/pm-skills repository, you can extend functionality by adding new skill files or updating existing ones in the skills/ directory. Claude automatically picks up changes on the next plugin load without requiring code compilation or deployment.

Summary

• The pm-product-discovery plugin provides slash commands for structured product management workflows in Claude.
• All capabilities are defined in markdown files: the manifest at .claude-plugin/plugin.json, skills in skills/*/SKILL.md, and command mappings in commands/*.md.
• Key commands include /pm-product-discovery:interview for transcript analysis and /pm-product-discovery:discover for full discovery cycles.
• Skills execute step-by-step frameworks (ICE, RICE, JTBD) and return structured markdown deliverables.
• The plugin is extensible by editing markdown files in the phuryn/pm-skills repository.

Frequently Asked Questions

What file format does the pm-product-discovery plugin use for its skills?

The plugin uses markdown files with YAML frontmatter headers. Each skill is a self-contained .md file containing the description, instructions, and templates. This design allows version control through Git and easy editing without programming knowledge.

Do I need to install the pm-product-discovery plugin separately?

No. According to the phuryn/pm-skills repository structure, Claude loads the plugin automatically when it detects the .claude-plugin/plugin.json manifest. You simply type the slash commands (e.g., /pm-product-discovery:interview) in a Claude chat that has access to this repository.

Can I modify the prioritize-features skill to use a different framework?

Yes. Because skills are markdown-based, you can edit prioritize-features/SKILL.md in the repository to adjust the evaluation criteria (for example, switching from ICE to RICE scoring) or modify the output table format. Changes take effect on the next Claude session.

Where can I find the complete list of available commands?

The authoritative list resides in pm-product-discovery/README.md in the repository root. This file catalogs all skills and commands, including newer additions like analyze-feature-requests (accessed via /pm-product-discovery:triage-requests) and their typical input/output formats.

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