Traction scoreGitHub stars can be faked, so popularity alone can be misleading. Traction Score looks for broader signs of recent attention, adoption, and active maintenance.
This repository provides a comprehensive set of plug-and-play skills designed to standardize how AI agents handle product management. By implementing a Triple Diamond framework, it helps AI product managers move from foundational discovery through to tool-specific execution with consistent, contract-enforced outputs.
The collection includes specialized workflows and sub-agents that transform vague product requests into structured deliverables, supported by a library of output samples and CI-enforced contracts to ensure reliability across the product lifecycle.
WHO IT'S FOR
AI product managers
standardize AI agent workflows using triple diamond
AI agent / automation builders
implement structured PM capabilities into AI agents
Compatible AgentsThe repository documents support for these agents. The skills may also work with other agents that can load SKILL.md files, but they may need some setup or small changes.
Follow the documented setup, then try a first task.
I’m using
Claude Code marketplace
Installs: The complete pm-skills plugin: all 68 skills, 10 workflow commands, the /chain runner, templates, sub-agents, and sample outputs. · Claude Code
Before you start
Claude Code with plugin marketplace support
Platform: Claude Code
Enter these slash commands in a Claude Code agent session, in the shown order; they are not terminal commands.
Installs: All 68 skills from the repository's skills/ directory. The full library preserves seamless handoffs between related skills. · Claude Code, Cursor, GitHub Copilot, Cline, other skills-ecosystem agents
Before you start
npx available
Platform: Cross-agent skills ecosystem
Run this command in a terminal. It installs the complete pm-skills library into the agent's default skills directory.
Installs: The complete repository, including the full skill library and its accompanying files. · GitHub Copilot, Cursor, Windsurf, Codex, OpenCode, Cline, Continue
Before you start
Git available
Platform: Platforms that load or auto-discover skills from a repository clone
In a terminal, change to the parent directory where the new pm-skills repository directory should be created, then run the documented clone command.
Installs: The pm-skills repository's skill set, installed and synchronized into the user's environment for compatible assistants to auto-discover. · OpenSkills-compatible AI assistants
Before you start
npm available
Platform: OpenSkills-compatible environments
Run this command in a terminal to install OpenSkills globally.
npm install -g openskills
Run this command in a terminal to install the pm-skills repository through OpenSkills.
openskills install product-on-purpose/pm-skills
Run this command in a terminal to synchronize the installed skills into the environment.
Use the define-problem-statement skill to create a concise problem statement from the context below. Identify the affected user segment, pain points and supporting evidence, business impact, measurable success criteria with baselines and targets, constraints, and open questions. Keep the framing focused on the problem rather than prescribing a solution.
Context: [initiative or situation]
User segment: [specific users]
Evidence: [research, support tickets, or behavioral data]
Business impact: [revenue, retention, growth, or strategic effect]
Known constraints: [technical, resource, regulatory, or dependency constraints]
Use the develop-solution-brief skill to draft a one-page solution brief for the problem and proposed approach below. Include a 2–3 sentence problem recap, a non-technical solution description, 3–5 essential features, measurable success targets, explicit trade-offs, the largest risks and mitigations, and 3–5 immediate next steps with owners where known.
Problem: [agreed problem statement]
Proposed approach: [solution idea]
Success measures: [metrics and targets]
Constraints or exclusions: [scope limits]
Known risks: [risks]
Use the deliver-release-notes skill to turn the shipped changes below into concise, user-facing draft release notes. Select 1–3 highlights, translate technical changes into user benefits, group the rest under New Features, Improvements, and Bug Fixes, and disclose any known issues. Keep each item to 1–2 sentences and do not include internal ticket IDs or code names.
Audience: [users or customers]
Shipped changes: [engineering changelog, completed tickets, or pull-request summaries]
Known issues: [limitations or none]
Optional coming-soon item: [item or omit]