What Level of Detail Is Provided for Each Skill in the PM Skills Repository?

Each skill in the phuryn/pm-skills repository is defined as a self-contained markdown file that provides explicit input contracts, step-by-step execution workflows, domain-specific evaluation frameworks, and authoritative reference links, enabling AI agents to execute complex product management tasks deterministically without external context.

The PM Skills Marketplace organizes every skill as a standalone SKILL.md file following a strict, information-rich template. This standardized approach ensures that every skill definition contains sufficient detail to function as both documentation and executable specification.

The Standardized SKILL.md Template Structure

Every skill file in the repository follows a consistent schema that breaks down into several high-fidelity sections:

YAML Header and Discovery Metadata

The file begins with a YAML frontmatter block containing the name and a concise description used for discovery and indexing. For example, in pm-toolkit/skills/review-resume/SKILL.md, the header provides machine-readable identifiers that allow the system to locate and invoke the correct skill.

Human-Readable Context and Purpose

Each skill includes a human-readable title and a Purpose & Context paragraph that tells the AI model why the skill exists and when to invoke it. This section typically contains role-playing instructions such as "You are an expert resume reviewer…" to establish the persona and expertise level required.

Input Arguments and Response Structure

The level of detail extends to explicit interfaces:

• Input Arguments: A declared list of variables (e.g., $RESUME, $JOB_POSTING) that the skill expects, with type definitions and descriptions
• Response Structure: A step-by-step outline (often numbered) guiding the model on how to format its output, such as Sections 1️⃣ Introduction, 2️⃣ Detailed Feedback on 10 Best Practices, 3️⃣ Conclusion

Domain-Specific Content and Evaluation Criteria

The core of each skill contains executable domain knowledge that varies by function:

• Evaluation criteria (what specific attributes to assess)
• Guidance (how to improve or optimize)
• Examples (good vs. weak samples for comparison)

For instance, the review-resume skill contains the complete "10 Best Practices for PM Resumes" checklist, while pm-product-strategy/skills/swot-analysis/SKILL.md embeds the full SWOT matrix methodology with actionable recommendation frameworks.

Process Steps and Edge Cases

Ordered execution workflows specify deterministic actions the model should follow. The product-vision skill in pm-product-strategy/skills/product-vision/SKILL.md defines a 6-step process (review inputs, identify problem, envision future state, etc.). Additionally, Notes / Edge Cases sections provide nuanced advice for handling ambiguous inputs or special conditions, such as "If $JOB_POSTING is provided, scan resume for keywords…"

Additional Guidelines and Authority Links

Each skill concludes with tone and style instructions (e.g., "Keep feedback casual yet professional") and a Further Reading section containing markdown links to authoritative external material. The swot-analysis skill, for example, ends with links to SWOT tutorials that anchor the skill in real-world best practices.

Explicit Execution Contracts

The level of detail creates an explicit contract between the user and the AI:

• Input specification: The skill tells the model exactly what variables to expect (e.g., $RESUME, $JOB_POSTING, $MARKET_SEGMENT)
• Output formatting: Step-by-step instructions give the model a deterministic workflow that reduces hallucinatory output
• Boundary conditions: Clear guidance on handling missing inputs or edge cases

Embedded Domain Expertise

Rather than requiring external APIs or code execution, each skill file embeds complete frameworks:

• Product Strategy: The 9-section canvas in strategy skills covering Vision, Business Model, Pricing, etc.
• Resume Review: The 10-point checklist evaluating structure, metrics, and narrative flow
• SWOT Analysis: The full matrix with internal/external factor evaluation criteria
• User Stories: Acceptance criteria templates and INVEST framework applications

This embedded expertise ensures the AI can perform complex product management tasks—such as resume review, vision crafting, or competitive analysis—using only the markdown file contents.

Practical Usage and File Locations

The detailed specifications enable simple slash-command invocation. When a user types:

/review-resume

The system loads pm-toolkit/skills/review-resume/SKILL.md, prompts for the $RESUME input, and executes the 10-best-practice checklist workflow defined in the file.

Similarly:

/swot-analysis — AI-code-review tool targeting developers

Executes the complete SWOT methodology stored in pm-product-strategy/skills/swot-analysis/SKILL.md, producing a full matrix with actionable recommendations.

Key skill files demonstrating this detailed structure include:

• pm-toolkit/skills/review-resume/SKILL.md – Resume evaluation with 10-point checklist
• pm-toolkit/skills/draft-nda/SKILL.md – Legal document drafting parameters
• pm-product-strategy/skills/product-vision/SKILL.md – 6-step vision crafting process
• pm-product-strategy/skills/swot-analysis/SKILL.md – Matrix evaluation framework
• pm-product-discovery/skills/brainstorm-ideas-new/SKILL.md – Three-perspective ideation workflow
• pm-market-research/skills/competitor-analysis/SKILL.md – Competitive landscape evaluation
• pm-execution/skills/user-stories/SKILL.md – Acceptance criteria specifications
• pm-data-analytics/skills/sql-queries/SKILL.md – Query structure templates
• pm-go-to-market/skills/gtm-strategy/SKILL.md – Launch planning methodology
• pm-ai-shipping/skills/shipping-artifacts/SKILL.md – Release documentation standards

Summary

The level of detail in the PM Skills repository is uniformly high across all skill definitions:

• Self-contained execution: Each SKILL.md file contains all necessary context, eliminating dependency on external data
• Deterministic workflows: Step-by-step process instructions ensure consistent output quality
• Rich domain frameworks: Complete evaluation criteria and methodologies embedded directly in markdown
• Extensible templating: New skills inherit the detailed structure by copying the standardized template, ensuring immediate AI compatibility

Frequently Asked Questions

How does the PM Skills repository ensure consistent detail across different skill types?

The repository enforces a strict template structure that every SKILL.md file must follow, including mandatory sections for YAML headers, input arguments, response structure, and domain-specific content. This standardization ensures that whether the skill handles resume review (pm-toolkit/skills/review-resume/SKILL.md) or SWOT analysis (pm-product-strategy/skills/swot-analysis/SKILL.md), the AI receives the same depth of instruction and context.

What specific information is included in the input arguments section?

Each skill defines an explicit variable list using the $VARIABLE_NAME convention, accompanied by type descriptions and usage instructions. For example, the resume review skill specifies $RESUME (the content to review) and optionally $JOB_POSTING (for keyword alignment), while strategy skills might define $MARKET_SEGMENT or $COMPETITOR_LIST, creating a clear contract for data exchange.

Can the AI execute these skills without additional code or APIs?

Yes. The markdown files act as executable specifications containing all necessary domain knowledge, evaluation frameworks, and process steps. When a user invokes a slash command like /brainstorm-ideas-new, the AI reads the corresponding SKILL.md file and follows the prescribed workflow—including multi-step processes like generating three perspectives (PM, Designer, Engineer) and prioritizing the top five ideas—without requiring external scripts or API calls.

How does the repository handle edge cases and ambiguous inputs?

Each skill includes a Notes / Edge Cases section that provides specific guidance for handling special conditions, missing variables, or ambiguous inputs. For instance, the resume review skill instructs the AI on how to proceed if $JOB_POSTING is omitted versus when it is provided, ensuring robust execution across variable input quality.

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