How to Use the pm-execution Plugin for Daily Product Management Tasks

The pm-execution plugin transforms Claude into a structured product management assistant by parsing markdown-based skills and commands to generate user stories, OKRs, sprint plans, and retrospectives through simple slash commands.

The pm-execution plugin, located in the phuryn/pm-skills repository, provides a Claude-compatible extension that operationalizes product management frameworks through declarative markdown files. By invoking slash commands, product managers can instantly generate structured artifacts like user stories, sprint plans, and OKRs without switching contexts. This guide demonstrates how to leverage the plugin's architecture and command syntax for your daily PM workflow.

Plugin Architecture: Skills, Commands, and Manifest

The pm-execution plugin follows the Claude plugin convention, storing all logic in static markdown files for safety and portability. The architecture consists of three core components:

Plugin Manifest

The file pm-execution/.claude-plugin/plugin.json defines the plugin's metadata, version, and keyword taxonomy. This JSON manifest registers the plugin with the Claude engine and maps invocation keywords to their respective command handlers.

Skills Directory

Skills are declarative markdown specifications that describe what each capability does and the step-by-step process it follows. When you invoke a command, the engine loads the corresponding skill file and executes its defined workflow. Key skill files include:

• pm-execution/skills/user-stories/SKILL.md: Implements the 3 C's (Card, Conversation, Confirmation) and INVEST criteria for comprehensive backlog items.
• pm-execution/skills/wwas/SKILL.md: Defines the Why-What-Acceptance format for strategic backlog items that embed business context.
• pm-execution/skills/sprint-plan/SKILL.md: Handles capacity estimation, velocity tracking, and story selection workflows for iteration planning.

Commands Directory

Commands serve as the user-facing entry points. Each command file declares an invocation syntax, workflow steps, and the specific skill it delegates to for processing:

• pm-execution/commands/write-stories.md: Entry point for generating user stories, job stories, and WWA items.
• pm-execution/commands/plan-okrs.md: Generates OKR sets aligned to company-level goals.
• pm-execution/commands/sprint.md: Orchestrates sprint planning, retrospectives, and release note generation.
• pm-execution/README.md: Provides a high-level overview of all available skills and commands.

Daily Workflow Commands and Examples

When you invoke a command like /write-stories user ..., the plugin engine parses the arguments, loads the corresponding skill markdown file (e.g., pm-execution/skills/user-stories/SKILL.md), and executes the step-by-step process described within it. The engine replaces placeholders such as $PRODUCT and $FEATURE with your supplied context, returning formatted markdown ready for insertion into tickets, Confluence pages, or version control documentation.

Generating User Stories

Use the /write-stories command with three distinct formats to match your planning methodology:

User Stories (3 C's + INVEST criteria):

/write-stories user Allow users to export reports as PDF and CSV

This invokes the user-stories skill to output titled stories with descriptions, design link placeholders, and 4-6 acceptance criteria per item.

Job Stories (Jobs-to-be-Done format):

/write-stories job Notification system for task deadlines

Produces job-story formatted items following the "When..., I want..., so I can..." structure with contextual acceptance criteria.

WWA Items (Strategic context):

/write-stories wwa Dark mode for the mobile app

Generates Why-What-Acceptance items that embed strategic "why" reasoning before technical specifications, using the workflow defined in pm-execution/skills/wwas/SKILL.md.

Planning OKRs

Align team objectives with company goals using the OKR planning skill:

/plan-okrs Growth team Q2 — company goal is 50% ARR increase

This command processes the plan-okrs skill to return three OKR sets, each containing a qualitative objective, 3-5 measurable key results, and an alignment map based on the input context.

Sprint Lifecycle Management

The /sprint command handles the complete sprint lifecycle through subcommands:

Sprint Planning:

/sprint plan 2-week sprint, 4 engineers, focus on checkout improvements

Outputs capacity breakdowns, recommended story selection, risk identification, and a complete sprint plan markdown based on the sprint-plan skill logic in pm-execution/skills/sprint-plan/SKILL.md.

Retrospectives:

/sprint retro

After you provide team feedback, the plugin asks for the retro format (e.g., Start/Stop/Continue), then outputs grouped themes, key insights, and actionable items using the retrospective skill.

Release Notes:

/sprint release-notes

Generates a user-focused release notes document with sections for Highlights, New Features, Improvements, and Bug Fixes, ready for distribution to stakeholders.

Summary

• The pm-execution plugin in phuryn/pm-skills adds product management capabilities to Claude through markdown-based skills and commands.
• Three core components comprise the architecture: the plugin manifest (pm-execution/.claude-plugin/plugin.json), skill definitions (e.g., pm-execution/skills/user-stories/SKILL.md), and command entry points (e.g., pm-execution/commands/write-stories.md).
• Daily tasks map to specific slash commands: /write-stories for backlog generation, /plan-okrs for objective setting, and /sprint for iteration management.
• All logic resides in static markdown, making the plugin language-agnostic, extensible, and safe for proprietary environments.

Frequently Asked Questions

What is the pm-execution plugin?

The pm-execution plugin is a Claude-compatible extension that provides structured product management commands. It reads declarative markdown skill files from the phuryn/pm-skills repository and generates standardized PM artifacts like user stories, OKRs, and sprint plans through natural language slash commands.

How do I install the pm-execution plugin?

Install the plugin by placing the pm-execution directory from the phuryn/pm-skills repository into your Claude plugins folder. The plugin self-registers via pm-execution/.claude-plugin/plugin.json without requiring additional dependencies or build steps.

Can I customize the output format of generated stories?

Yes. Since output templates live in skill markdown files (e.g., pm-execution/skills/user-stories/SKILL.md), you can modify the structure, acceptance criteria prompts, or formatting conventions by editing these files. Changes take effect immediately on the next command invocation without restarting the Claude environment.

Is the pm-execution plugin secure to use with proprietary data?

Yes. The plugin operates entirely through static markdown parsing and does not execute arbitrary code or external API calls. All processing happens within the Claude environment using your provided context, making it safe for internal product details and proprietary roadmaps.

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