How the PM Skills Marketplace Encodes Product Management Frameworks like Teresa Torres and Marty Cagan

The PM Skills Marketplace transforms product management frameworks into executable markdown skills, embedding Teresa Torres' Continuous Discovery Habits and Marty Cagan's INSPIRED methodology directly into AI-ready SKILL.md files with structured front-matter and step-by-step workflows.

The phuryn/pm-skills repository provides a structured approach to encoding product management frameworks as machine-readable instructions. By packaging expert methodologies into self-contained markdown units, the marketplace enables AI assistants to execute complex product discovery and strategy workflows without requiring external lookups.

What Is a Skill in the PM Skills Marketplace?

Each skill is a self-contained knowledge unit implemented as a markdown file named SKILL.md. These files live within plugin directories (such as pm-product-discovery or pm-product-strategy) and contain three critical components:

1. Front-matter with a concise name and description that explicitly cites the source framework
2. Detailed instructions that guide the assistant through the steps, terminology, and decision-rules defined by that methodology
3. References including links, tables, and formulas that let the model surface original concepts

When a user query matches a skill's trigger phrase (for example, "Help me build an Opportunity Solution Tree"), Claude automatically loads the corresponding SKILL.md file and follows its embedded workflow.

How Teresa Torres' Continuous Discovery Habits Is Encoded

The repository implements Torres' methodology through a dedicated skill file that translates her book into an executable decision tree.

File Structure and Front-Matter

In pm-product-discovery/skills/opportunity-solution-tree/SKILL.md, the skill description begins with an explicit citation: "Build an Opportunity Solution Tree (OST)... Based on Teresa Torres' Continuous Discovery Habits."

The front-matter establishes the framework's authority while the body encodes Torres' specific principles, including "One outcome at a time" and the distinction between "Opportunities, not features."

The Opportunity Solution Tree Implementation

The skill defines the four-level OST hierarchy that mirrors Torres' methodology:

• Outcome – The single measurable business result to improve
• Opportunities – 3-5 user needs or pain points derived from research
• Solutions – Multiple product ideas generated through product-trio brainstorming
• Experiments – Fast validation tests for the top solutions

Example Usage

When a user requests continuous discovery, the assistant loads the skill and executes the encoded workflow:

User: "We want to improve onboarding retention for our SaaS product. 
Build an Opportunity Solution Tree."

Assistant (Claude) – loads opportunity-solution-tree skill:

1. Define the desired outcome – "Increase 7-day retention to 40%"
2. Map 3-5 opportunities from user research
3. Generate ≥3 solutions per opportunity (product-trio brainstorming)
4. Propose fast experiments for the top solutions

How Marty Cagan's INSPIRED Framework Is Encoded

Marty Cagan's product strategy methodology from INSPIRED and TRANSFORMED is embedded through citation in documentation and implementation in the strategy plugin.

Framework Citation in Documentation

The README.md file lists "Marty Cagan – INSPIRED and TRANSFORMED" as a core source framework, establishing the intellectual foundation for the marketplace's approach to product strategy.

The Product Strategy Skill Structure

In pm-product-strategy/skills/product-strategy/SKILL.md, the skill implements a step-by-step canvas that mirrors Cagan's prescribed flow:

1. Inspire – Describe the problem and opportunity
2. Target – Define the ideal customer segment (ICP)
3. Rationale – Articulate market timing and strategic fit
4. Idea – Outline the solution concept
5. Solution – Detail MVP features and scope
6. Validate – Suggest early validation experiments

Example Usage

The skill triggers when users reference Cagan's framework:

User: "Create a product strategy for a new AI-writing assistant, 
using the INSPIRED framework."

Assistant – loads product-strategy skill:

1. Inspire – Describe the problem: "Writers struggle with consistency"
2. Target – Define ICP: "Content marketers at B2B SaaS companies"
3. Rationale – "AI capability maturation meets remote work content demands"
4. Idea – "AI co-pilot that learns brand voice"
5. Solution – "Browser extension with style-matching algorithms"
6. Validate – "Landing page tests with 10 pilot customers"

The Technical Mechanism: From Markdown to AI Execution

The encoding system relies on commands that chain multiple skills into end-to-end flows. For example, the /discover command in the pm-product-discovery plugin chains the opportunity-solution-tree skill with other discovery skills, delivering a full continuous-discovery process built on Torres' methodology.

Because the source framework is baked directly into the markdown, the assistant can:

• Cite original authors without external searches
• Apply framework-specific heuristics (e.g., "One outcome at a time")
• Maintain consistency with the original authors' intent

The repository's structure ensures that SKILL.md files serve as the single source of truth, with the README.md files in each plugin directory (pm-product-discovery/README.md and pm-product-strategy/README.md) explaining how skills group into commands.

Summary

• Skills are markdown files named SKILL.md that package frameworks into executable instructions with front-matter citations.
• Teresa Torres' OST lives in pm-product-discovery/skills/opportunity-solution-tree/SKILL.md and encodes the four-level tree (outcome → opportunities → solutions → experiments).
• Marty Cagan's INSPIRED methodology is cited in the root README.md and implemented in pm-product-strategy/skills/product-strategy/SKILL.md as a six-step canvas.
• Commands chain skills into workflows, allowing the /discover command to execute Torres' continuous discovery process end-to-end.
• Self-contained citations ensure AI assistants reference original sources without runtime web searches.

Frequently Asked Questions

What file format does PM Skills Marketplace use to encode frameworks?

The repository uses markdown files named SKILL.md to encode each framework. These files contain YAML front-matter for metadata and structured instructions that guide AI assistants through the specific methodology, eliminating the need for external API calls during execution.

How does the Opportunity Solution Tree skill implement Teresa Torres' methodology?

The skill file at pm-product-discovery/skills/opportunity-solution-tree/SKILL.md explicitly cites Torres' Continuous Discovery Habits in its description. It encodes her four-level OST structure and embeds her core principles, such as focusing on "one outcome at a time" and distinguishing opportunities from features, directly into the instruction set.

Can I combine multiple framework skills into a single workflow?

Yes. The marketplace uses commands to chain skills into integrated workflows. For example, the /discover command combines the opportunity-solution-tree skill with other discovery skills, allowing an AI assistant to execute a complete continuous-discovery process that adheres to Torres' methodology while potentially incorporating other complementary frameworks.

Where are the source citations for frameworks like Marty Cagan's INSPIRED located?

Citations appear in two locations: the root README.md lists Cagan's INSPIRED and TRANSFORMED books as core sources, while the individual SKILL.md files in the pm-product-strategy plugin contain the detailed implementation steps and specific references to Cagan's product-strategy canvas methodology.

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