# Main Features of the PM-Skills Project: An AI-Powered Product Management Marketplace

> Explore the main features of the PM-Skills project, an AI product management OS with 9 plugins, 68 skills, and 42 commands for the complete product lifecycle.

- Repository: [Pawel Huryn/pm-skills](https://github.com/phuryn/pm-skills)
- Tags: getting-started
- Published: 2026-07-03

---

**The pm-skills project is an AI-operated product management operating system that ships nine specialized plugins for Claude, containing 68 reusable skills and 42 slash commands covering the entire product lifecycle from discovery to shipping.**

The **pm-skills** repository by phuryn provides an open-source marketplace of AI-native plugins designed to transform Claude (Code, Cowork) and other LLM assistants into comprehensive product management tools. This modular system encodes proven PM frameworks into self-contained markdown skills and workflow commands, enabling offline-friendly execution of complex product workflows without external API calls.

## Modular Plugin Architecture Covering the Full Product Lifecycle

The project organizes functionality into nine distinct plugins, each targeting a specific phase of product management. According to the repository structure in [`README.md`](https://github.com/phuryn/pm-skills/blob/main/README.md), these plugins are self-contained directories that can be installed independently via `claude plugin install <plugin>@pm-skills` or collectively as the complete marketplace.

### The Nine Core Plugins

| Plugin | Purpose | Skills | Commands |
|--------|---------|--------|----------|
| **pm-toolkit** | PM utilities (resume, legal docs) | 4 | 5 |
| **pm-product-discovery** | Ideation and assumption mapping | 13 | 5 |
| **pm-product-strategy** | Vision and business models | 12 | 5 |
| **pm-execution** | PRDs, OKRs, roadmaps | 16 | 11 |
| **pm-market-research** | Personas and journey maps | 7 | 3 |
| **pm-data-analytics** | SQL and cohort analysis | 3 | 3 |
| **pm-go-to-market** | ICP and growth loops | 6 | 3 |
| **pm-marketing-growth** | Positioning and metrics | 5 | 2 |
| **pm-ai-shipping** | AI codebase documentation | 2 | 5 |

### Plugin Registration and Structure

Each plugin declares its metadata through a [`.claude-plugin/plugin.json`](https://github.com/phuryn/pm-skills/blob/main/.claude-plugin/plugin.json) manifest file. For example, [`pm-toolkit/.claude-plugin/plugin.json`](https://github.com/phuryn/pm-skills/blob/main/pm-toolkit/.claude-plugin/plugin.json) registers the toolkit plugin with its version and capabilities, while [`pm-product-strategy/.claude-plugin/plugin.json`](https://github.com/phuryn/pm-skills/blob/main/pm-product-strategy/.claude-plugin/plugin.json) handles strategy-focused functionality. This declarative approach allows the Claude ecosystem to automatically discover and load plugin metadata.

## Skill-First Design and Command Orchestration

The architecture separates domain knowledge (**skills**) from workflow logic (**commands**), creating reusable components that can be chained together for complex operations.

### Self-Contained Skill Files

Skills are plain markdown files stored under `<plugin>/skills/` directories. For instance, [`pm-toolkit/skills/review-resume/SKILL.md`](https://github.com/phuryn/pm-skills/blob/main/pm-toolkit/skills/review-resume/SKILL.md) encodes the XYZ+S resume formula, while skills in `pm-product-discovery/skills/` contain frameworks like Opportunity Solution Trees and assumption mapping techniques. These markdown files follow a universal schema that makes them portable across different AI assistants including Gemini, OpenCode, and Cursor.

### Workflow Commands

Commands are orchestration scripts stored in `<plugin>/commands/` directories. The file [`pm-execution/commands/write-prd.md`](https://github.com/phuryn/pm-skills/blob/main/pm-execution/commands/write-prd.md) defines the `/write-prd` slash command, which chains multiple skills to guide users through creating an 8-section Product Requirements Document. When invoked, these commands automatically assemble the appropriate skills and guide conversations step-by-step.

## AI-Native Interaction and Cross-Platform Compatibility

The system is designed specifically for LLM consumption, loading knowledge on-the-fly without requiring external API calls. This architecture makes the **pm-skills** project completely offline-friendly and privacy-preserving, as all processing happens within the Claude environment.

While optimized for Claude, the markdown-based skill format is cross-compatible. Skills can be dropped into other LLM environments like Gemini’s `.gemini/skills/` or OpenCode’s `.opencode/skills/` directories, maintaining functionality across different AI assistants.

## Practical Usage Examples

The marketplace exposes functionality through intuitive slash commands that assemble complex workflows instantly.

To conduct full product discovery:

```text
/discover AI-powered meeting summarizer for remote teams

```

This chains skills from `pm-product-discovery` including **brainstorm-ideas-new**, **identify-assumptions-new**, **prioritize-assumptions**, and **brainstorm-experiments-new**.

To write a comprehensive PRD:

```text
/write-prd Smart notification system that reduces alert fatigue

```

This invokes the **create-prd** skill from [`pm-execution/commands/write-prd.md`](https://github.com/phuryn/pm-skills/blob/main/pm-execution/commands/write-prd.md), producing an 8-section PRD template.

To generate SQL analytics:

```text
/write-query Show me monthly active users by country for Q4 2025 (BigQuery)

```

This uses the **sql-queries** skill from the `pm-data-analytics` plugin to translate natural language into executable SQL.

To audit AI-generated code:

```text
/ship-check the payments service

```

This runs **shipping-artifacts** and **intended-vs-implemented** skills from `pm-ai-shipping` to produce documentation and static security audits.

## Summary

- The **pm-skills** project provides nine specialized plugins containing 68 skills and 42 commands
- Skills are self-contained markdown files encoding proven PM frameworks like SWOT, Opportunity Solution Trees, and Lean Canvas
- Commands orchestrate workflows by chaining skills through slash commands like `/discover`, `/strategy`, and `/write-prd`
- The [`.claude-plugin/plugin.json`](https://github.com/phuryn/pm-skills/blob/main/.claude-plugin/plugin.json) manifest system enables automatic plugin discovery and registration
- Cross-platform markdown format works with Claude, Gemini, OpenCode, and other LLM assistants
- Completely offline-capable and privacy-preserving with no external API dependencies

## Frequently Asked Questions

### What is the pm-skills project?

The pm-skills project is an open-source marketplace of AI-native plugins that transform LLM assistants into product management operating systems. It ships nine plugins containing markdown-based skills and commands covering the entire product lifecycle from ideation to shipping documentation.

### How do I install and use pm-skills plugins?

You can install individual plugins via `claude plugin install <plugin>@pm-skills` or install the entire marketplace at once. Once installed, invoke workflows using slash commands like `/discover` or `/write-prd`, which automatically load the relevant skills from directories like `pm-product-discovery/skills/` or `pm-execution/skills/`.

### Can I use pm-skills with assistants other than Claude?

Yes, the skills are stored as plain markdown files following a universal schema, making them compatible with Gemini, OpenCode, Cursor, and other LLM assistants. Simply copy the skill files from directories like `<plugin>/skills/` into your assistant's skills directory.

### What types of product management tasks does pm-skills cover?

The project covers the complete product lifecycle including product discovery (assumption mapping, experiments), strategy (business models, competitive analysis), execution (PRDs, OKRs, roadmaps), market research (personas, journey maps), data analytics (SQL generation, A/B testing), go-to-market planning (ICP, messaging), and AI shipping documentation.