Main Features of phuryn/pm-skills: A Modular AI Marketplace for Product Managers

The phuryn/pm-skills repository is a collection of 68+ AI-powered product management frameworks organized into nine specialized plugins, offering reusable skills and slash commands that automate the entire product lifecycle from discovery to launch.

The phuryn/pm-skills repository functions as an open-source marketplace where product managers access structured AI capabilities through a universal markdown format. Each functional area of product management is packaged as a standalone plugin containing specialized skills and executable commands. This modular architecture allows AI assistants like Claude, Gemini, Cursor, and OpenCode to ingest proven PM frameworks and execute complex workflows through simple slash commands.

Core Architecture: Skills, Commands, and Plugins

The repository organizes functionality into three distinct layers that work together to provide end-to-end automation.

Skills: Reusable Knowledge Bases

Skills are markdown-based frameworks stored in <plugin>/skills/<skill-name>/SKILL.md files. Each skill encodes proven methodologies—such as Opportunity Solution Trees, SWOT analysis, or PRD templates—into a structured schema that AI assistants can parse and execute.

For example, the brainstorm-ideas-new skill resides in pm-product-discovery/skills/brainstorm-ideas-new/SKILL.md and contains the complete prompt engineering required to generate structured product ideas. Skills follow a strict markdown schema defining inputs, outputs, and optional examples, making them portable across different AI platforms.

Commands: End-to-End Workflows

Commands are orchestration files located in <plugin>/commands/<command-name>.md that chain multiple skills into cohesive workflows. Unlike individual skills, commands handle sequencing, context passing, and user interaction templates.

When you invoke /discover, the command file at pm-product-discovery/commands/discover.md automatically executes a sequence of skills: brainstorm-ideas-new, identify-assumptions-new, prioritize-assumptions, and brainstorm-experiments-new. This abstraction allows product managers to run complete discovery cycles without manually managing each step.

Plugin Manifest Structure

The root README.md serves as the plugin catalog, enumerating all nine functional areas. When installing via the marketplace command claude plugin marketplace add phuryn/pm-skills, the installer reads this manifest and auto-installs every plugin folder, making all skills and commands immediately available.

The Nine Plugin Categories

The repository organizes 68+ skills across nine specialized plugins, each targeting a specific domain of product management.

pm-product-discovery

This plugin handles ideation, assumption mapping, and experiment design. Key skills include opportunity-solution-tree, metrics-dashboard, and brainstorm-ideas-new. Representative commands such as /discover, /brainstorm, and /triage-requests guide teams through the ambiguous early stages of product development.

pm-product-strategy

Focused on vision and business models, this plugin contains skills like product-strategy, lean-canvas, swot-analysis, and pricing-strategy. The /strategy command generates a nine-section canvas, while /business-model and /value-proposition help articulate market positioning.

pm-execution

The operational core for day-to-day PM deliverables includes skills for create-prd, user-stories, test-scenarios, and prioritization-frameworks. Commands like /write-prd, /plan-okrs, /sprint, and /meeting-notes transform high-level ideas into structured documentation and sprint plans.

pm-market-research

This plugin supports user research and competitive analysis through skills such as user-personas, market-sizing, competitor-analysis, and sentiment-analysis. Commands including /research-users, /competitive-analysis, and /analyze-feedback systematize qualitative and quantitative research workflows.

pm-data-analytics

Data-driven PMs utilize skills like sql-queries, cohort-analysis, and ab-test-analysis. The /write-query skill emits parameterized BigQuery statements, while /analyze-cohorts and /analyze-test provide statistical frameworks for interpreting user behavior.

pm-go-to-market

Covering launch strategy and growth mechanics, this plugin includes gtm-strategy, ideal-customer-profile, growth-loops, and competitive-battlecard. Commands such as /plan-launch, /growth-strategy, and /battlecard prepare teams for market entry.

pm-marketing-growth

Focused on positioning and metrics, skills include marketing-ideas, positioning-ideas, product-name, and north-star-metric. The /market-product and /north-star commands help teams identify growth levers and define success metrics.

pm-toolkit

Utility functions for professional PM tasks include review-resume, draft-nda, privacy-policy, and grammar-check. Commands like /review-resume, /draft-nda, and /proofread handle administrative and legal documentation.

pm-ai-shipping

A specialized plugin for documenting and auditing AI-generated codebases contains skills such as shipping-artifacts and intended-vs-implemented. Commands including /ship-check, /document-app, and /security-audit-static ensure compliance and documentation standards for AI-assisted development.

How It Works Under the Hood

The repository implements a declarative architecture that separates content from execution logic.

Skill Definition Schema

Each skill follows a universal markdown format defined in SKILL.md files. The schema includes:

• Title and description: Clear identification of the framework
• Inputs: Required and optional parameters
• Outputs: Expected deliverables and format
• Examples: Sample invocations and expected results

This structure allows any AI assistant that understands markdown to parse the skill definition without custom integration code.

Command Orchestration

Command files specify execution graphs using simple markdown frontmatter and content sections. When a user invokes a slash command, the AI assistant loads the corresponding .md file from the <plugin>/commands/ directory, reads the skill dependencies, and executes them in sequence while maintaining context between steps.

Auto-Loading Behavior

When installed, the repository enables two usage modes:

1. Automatic loading: The AI assistant detects PM-related tasks and suggests relevant skills from the installed plugins
2. Explicit invocation: Users can force-load specific skills using the /plugin-name:skill-name prefix syntax

Practical Usage Examples

The following slash commands demonstrate real-world usage patterns. These are self-contained workflows that execute immediately after installing the repository.

# Execute a complete discovery cycle

/discover AI-powered meeting summarizer for remote teams

This command runs the full discovery sequence: idea generation, assumption identification, assumption prioritization, and experiment brainstorming.

# Generate a strategic canvas

/strategy B2B project-management tool for agencies

Invokes the product-strategy skill to produce a nine-section strategic canvas ready for stakeholder review.

# Create a Product Requirements Document

/write-prd Smart notification system that reduces alert fatigue

Uses the create-prd skill to generate an eight-section PRD with user stories, acceptance criteria, and success metrics.

# Draft legal documentation

/draft-nda Create a mutual NDA for vendor evaluation

Leverages the draft-nda skill to output a ready-to-review legal document tailored to the specified context.

Key Files to Explore

Understanding the repository structure requires examining these specific files:

• README.md: The plugin catalog and installation instructions located at the repository root
• pm-product-discovery/skills/brainstorm-ideas-new/SKILL.md: Example skill definition showing the markdown schema structure
• pm-execution/commands/write-prd.md: Example command implementation demonstrating workflow orchestration
• pm-toolkit/README.md: Documentation for utility skills including resume review and legal document generation
• pm-data-analytics/README.md: Analytics-focused skills and SQL generation capabilities
• pm-ai-shipping/README.md: Documentation for AI codebase auditing and shipping compliance

Each plugin directory contains its own README.md detailing the specific skills and commands available within that functional area.

Summary

• phuryn/pm-skills provides 68+ AI-powered product management frameworks organized into nine specialized plugins covering discovery, strategy, execution, research, analytics, GTM, marketing, utilities, and AI shipping.
• The architecture separates skills (reusable knowledge bases in SKILL.md files) from commands (workflow orchestrators in <command-name>.md files), enabling modular composition.
• All frameworks use a universal markdown format compatible with Claude, Gemini, Cursor, Kiro, and OpenCode, ensuring portability across AI assistants.
• Installation via claude plugin marketplace add phuryn/pm-skills auto-loads all plugins, making slash commands like /discover, /write-prd, and /strategy immediately available.
• The repository covers the complete product lifecycle, from initial ideation through launch, growth, data analysis, and AI-code compliance auditing.

Frequently Asked Questions

What is the difference between a skill and a command in phuryn/pm-skills?

A skill is a reusable knowledge base stored in a SKILL.md file that encodes a specific framework or methodology, such as creating a SWOT analysis or writing SQL queries. A command is a workflow orchestrator stored in a .md file within the commands/ directory that chains multiple skills together to accomplish complex tasks. For example, the /discover command runs four different skills in sequence to complete a full discovery cycle.

Which AI assistants are compatible with phuryn/pm-skills?

The repository uses a universal markdown format that any AI assistant capable of parsing markdown can consume. According to the source code, the skills are explicitly compatible with Claude, Gemini, OpenCode, Cursor, and Kiro. The SKILL.md schema uses standard markdown structures for inputs, outputs, and examples, ensuring broad compatibility without proprietary integrations.

How do I install and use the phuryn/pm-skills repository?

Install the repository using the marketplace command claude plugin marketplace add phuryn/pm-skills, which reads the root README.md manifest and auto-installs all nine plugin folders. Once installed, you can invoke skills directly through slash commands like /write-prd or /strategy, or force-load specific skills using the /plugin-name:skill-name syntax. All commands are self-contained and execute immediately without additional configuration.

What is the SKILL.md format and why is it important?

The SKILL.md format is a standardized markdown schema used across all skills in the repository. It defines the skill's title, description, required inputs, expected outputs, and optional examples in a predictable structure. This format is critical because it allows AI assistants to parse and execute the skill without custom coding, making the frameworks portable across different AI platforms and ensuring consistent behavior regardless of which assistant processes the request.

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