Benefits of Using pm-product-discovery Skills for a New Product

pm-product-discovery skills provide reusable, Markdown-based templates that standardize early-stage discovery workflows, enabling teams to reduce bias, accelerate learning, and maintain a single source of truth for product decisions.

The phuryn/pm-skills repository offers a structured toolkit for product managers navigating the uncertainty of new product development. By implementing pm-product-discovery skills, teams can transform ad-hoc discovery activities into repeatable, data-driven processes that integrate directly with version control systems. These skills encode proven frameworks—such as JTBD interview summarization and ICE/RICE prioritization—into executable commands that generate auditable artefacts.

What Are pm-product-discovery Skills?

pm-product-discovery skills are modular, Markdown-based guides stored in pm-product-discovery/skills/*.md that encode specific product discovery techniques. Each skill file contains front-matter metadata (name, description) that allows the CLI to discover and present them as executable commands. Unlike static documentation, these skills actively process input files—such as interview transcripts or CSV backlogs—and output structured Markdown artefacts ready for team review.

Key Benefits for New Product Development

Consistent Interview Capture

The summarize-interview skill eliminates variability in how teams document customer research. Located at pm-product-discovery/skills/summarize-interview/SKILL.md, this skill provides a step-by-step template that extracts date, participants, problems, likes, and action items from raw transcripts. By standardizing interview summaries around Jobs-to-be-Done (JTBD) principles, the skill ensures that qualitative data remains readable and actionable for stakeholders who did not attend the session.

Data-Driven Feature Prioritization

New products face relentless scope pressure. The prioritize-features skill, defined in pm-product-discovery/skills/prioritize-features/SKILL.md, applies ICE (Impact, Confidence, Ease) and RICE (Reach, Impact, Confidence, Effort) formulas alongside Opportunity Scoring to rank backlog items objectively. The skill evaluates impact, effort, risk, and strategic alignment, then outputs a ranked table of the top five features—removing subjective debates from roadmap planning.

Assumption Risk Management

Uncertainty kills new products. The prioritize-assumptions skill uses an Impact × Risk matrix combined with ICE scoring to visualize which assumptions threaten the product most. As implemented in pm-product-discovery/skills/prioritize-assumptions/SKILL.md, this skill flags high-risk, high-impact items and suggests lightweight experiments to validate them before committing engineering resources.

Version-Controlled Artefact Templates

Every pm-product-discovery skill generates Markdown outputs that can be version-controlled alongside source code. Unlike proprietary tools that lock data in inaccessible formats, these skills explicitly end with "Save as markdown" instructions, guaranteeing persistent artefacts. This architecture enables pull-request reviews for discovery decisions, ensuring that product strategy changes remain traceable and auditable through standard Git workflows.

Fast Onboarding for New Team Members

The repository ships with human-friendly CLI wrappers in pm-product-discovery/commands/*.md that eliminate the need to memorize complex frameworks. New product managers can run pm discover to instantly receive a guided workbook that orchestrates interview summarization and initial assumption capture. This command-line interface handles file I/O automatically, reducing the cognitive load required to execute professional-grade discovery.

Cross-Skill Integration

pm-product-discovery skills reference each other through "Domain Context" sections, creating a cohesive discovery funnel. For example, prioritize-features links to the prioritization-frameworks skill to ensure consistent terminology and calculation methods. This cross-referencing allows a single command to surface the full decision-making matrix without requiring users to manually navigate between disconnected documents.

How to Use pm-product-discovery Skills in Practice

The repository provides a unified CLI wrapper that invokes skills directly from the terminal. Below are minimal command-line snippets that process various input formats and emit structured Markdown artefacts.

Summarize a customer interview from a raw transcript:

pm interview --file transcript.txt --output interview-summary.md

Prioritize a feature backlog using ICE/RICE scoring:

pm prioritize-features --file features.csv --output top-features.md

Map assumptions and generate risk-mitigation experiments:

pm prioritize-assumptions --file assumptions.xlsx --output assumptions-matrix.md

Each command reads the supplied file (CSV, Excel, plain-text, or PDF), runs the associated skill logic, and emits a formatted Markdown file ready for sharing or committing to the repository.

Repository Structure and Key Files

Understanding the file layout helps teams customize skills for their specific discovery workflows:

• pm-product-discovery/skills/summarize-interview/SKILL.md – Defines the JTBD-focused interview summarization template and extraction logic.
• pm-product-discovery/skills/prioritize-features/SKILL.md – Implements ICE/RICE scoring and strategic alignment calculations for backlog ranking.
• pm-product-discovery/skills/prioritize-assumptions/SKILL.md – Contains the Impact × Risk matrix logic and experiment design framework.
• pm-product-discovery/commands/discover.md – Orchestrates the initial discovery workflow, invoking multiple skills in sequence.
• pm-product-discovery/commands/interview.md – CLI entry point that wraps the summarize-interview skill.
• pm-product-discovery/commands/triage-requests.md – Combines prioritize-features and prioritize-assumptions for backlog grooming sessions.

Summary

• pm-product-discovery skills standardize early-stage discovery through reusable Markdown templates stored in phuryn/pm-skills.
• Consistent documentation emerges from skills like summarize-interview, which structure qualitative data into JTBD-focused summaries.
• Objective prioritization results from ICE/RICE formulas applied by prioritize-features and prioritize-assumptions.
• Version control integration ensures all discovery artefacts remain auditable and shareable alongside code repositories.
• CLI automation reduces onboarding friction by wrapping complex frameworks into simple commands like pm discover and pm interview.

Frequently Asked Questions

What file formats do pm-product-discovery skills support?

The CLI commands accept CSV, Excel (.xlsx), plain text, and PDF inputs. The pm-toolkit core provides generic file I/O helpers that parse these formats and funnel content into the skill logic, emitting standardized Markdown outputs regardless of the input source.

How does pm-product-discovery integrate with existing Agile workflows?

Because skills generate Markdown artefacts, teams can store discovery outputs in the same Git repository as their codebase. This allows product decisions to undergo pull-request reviews alongside code changes, ensuring that discovery documentation remains current and traceable through standard version control practices.

Can organizations extend pm-product-discovery with custom skills?

Yes. Adding a new discovery method requires only creating an additional Markdown file in pm-product-discovery/skills/ with appropriate front-matter. The CLI automatically discovers all .md entries at runtime, making the toolkit extensible without modifying core logic. Organizations can maintain private forks or contribute new skills back to the repository.

Is pm-product-discovery suitable for non-technical product managers?

Yes. The CLI abstracts away implementation details, allowing users to run commands like pm interview without understanding the underlying Markdown structure or Git operations. The skills provide guided prompts and templates that walk users through professional discovery frameworks step-by-step, reducing the expertise required to execute rigorous product research.

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