A Claude Code workspace layer for product managers. Built on top of the Anthropic product-management plugin and WorkIQ for Microsoft 365, PM Studio adds personal context, voice matching, a knowledge base, quality standards, and a refinement agent pipeline to every command you run.
PM Studio uses a layered architecture:
- The plugin provides commands (
/product-management:*), deep methodology (6 skills), and MCP connectors. It auto-updates from the marketplace. - WorkIQ provides always-on access to your M365 data: emails, meeting transcripts, documents, and team discussions. Every command queries it by default.
- PM Studio provides the workspace: additional commands, agents, personal context, notes, voice matching, output organization, and quality standards (via
CLAUDE.md).
When you work inside the PM Studio workspace, all three layers are active. CLAUDE.md enhances everything.
Required:
- Claude Code (VS Code extension or CLI)
- Anthropic product-management plugin:
claude plugin marketplace add product-management@knowledge-work-plugins
- WorkIQ MCP for Microsoft 365 integration (emails, meetings, documents). Pre-configured in PM Studio's
.mcp.jsonand starts automatically. Accept the EULA on first use.
Optional:
- Additional MCP connectors (Slack, Linear, Figma, etc.) can be added to
.mcp.json - Handy.computer for voice-based brainstorming
WorkIQ is always on. Every command (plugin and PM Studio) queries WorkIQ for relevant emails, meetings, and documents by default. Tell Claude to skip it if you don't want it for a specific command.
git clone https://github.com/zoekdestep/pm-studio.gitOr download as ZIP from the GitHub repository page.
Open the pm-studio folder in VS Code, then open the Claude Code panel.
Fill in your context files:
| File | What to Add |
|---|---|
context/about-me.md |
Your role, stakeholders, priorities, PM philosophy |
context/product-context.md |
Product strategy, user segments, metrics, competitive landscape, technical constraints |
context/tone-and-voice.md |
Your writing style (good defaults included) |
context/examples/ |
4-5 of your best docs (these drive both voice AND structure for all output) |
Run any command. For example:
/product-management:write-spec
/strategy competitive-positioning
/shipping dark-mode
CLAUDE.md automatically enhances every command with your context, voice, and quality standards.
See NEXT-STEPS.md for a complete getting-started walkthrough.
| Command | Purpose | Output |
|---|---|---|
/strategy [topic] |
Write a research-backed strategy doc | output/strategies/ |
/shipping [feature] |
Write a shipping decision with metrics and recommendation | output/shipping-decisions/ |
/note [content] |
Capture a note to your knowledge base | notes/ |
/word [file] |
Export markdown to Word | output/ |
| Command | Purpose | Output |
|---|---|---|
/product-management:write-spec |
Write a feature spec or PRD | output/specs/ |
/product-management:roadmap-update |
Create or reprioritize a roadmap | output/roadmaps/ |
/product-management:stakeholder-update |
Generate a stakeholder update | output/stakeholder-updates/ |
/product-management:metrics-review |
Review and analyze product metrics | output/metrics-reviews/ |
/product-management:competitive-brief |
Create a competitive analysis brief | output/competitive-briefs/ |
/product-management:synthesize-research |
Synthesize user research into insights | output/research-syntheses/ |
All plugin commands are enhanced by PM Studio's context loading, voice matching, auto-save, and agent offering.
PM Studio's unique value: a post-generation refinement pipeline. After any document is generated (plugin or PM Studio), run agents to improve it.
| Agent | What It Does | Best For |
|---|---|---|
| reviewer | Two-pass review: structural completeness + Socratic reasoning analysis | All docs |
| editor | Reads your voice examples, applies full editorial pass | All docs before sharing |
| edge-case-finder | Surfaces scenarios, risks, and failure modes | Specs, roadmaps |
| metrics-designer | Suggests KPIs, guardrails, and measurement plans | Specs, shipping decisions |
| question-generator | Surfaces questions worth exploring | Strategy, research |
Common flow: generate a doc, run reviewer to find gaps, run editor for polish.
WorkIQ (always on): Every command queries WorkIQ for M365 data (emails, meetings, transcripts, documents) by default. No setup needed beyond accepting the EULA on first use.
Optional connectors use ~~category placeholders for additional data:
| Category | Data Source | Example MCP |
|---|---|---|
~~product analytics |
Usage metrics, trends | Your analytics MCP |
~~user feedback |
Support tickets, sentiment | Your feedback MCP |
~~project tracker |
Work items, rollout status | Linear, Jira MCPs |
Optional connectors are skipped silently when not available. See CONNECTORS.md for details.
pm-studio/
├── .claude/
│ ├── agents/ # 5 refinement agents
│ └── commands/ # PM Studio commands
├── context/
│ ├── examples/ # Your reference docs (voice + structure)
│ ├── about-me.md # Role, preferences, PM philosophy
│ ├── product-context.md # Strategy, metrics, competitors, tech constraints
│ └── tone-and-voice.md # Writing style guide
├── notes/
│ ├── ideas/ # Raw ideas, hunches
│ ├── research/ # User research, data findings
│ └── reading/ # Articles, books, learnings
├── output/ # Auto-saved documents (8 subdirectories)
├── .mcp.json # MCP server config (WorkIQ)
├── CLAUDE.md # Workspace enhancement layer
├── CONNECTORS.md # Connector architecture
└── NEXT-STEPS.md # Getting started guide
Voice: Put your best docs in context/examples/, edit context/tone-and-voice.md. The editor agent uses both.
Agents and commands: Add .md files to .claude/agents/ or .claude/commands/. Or ask Claude to create them for you.
Connectors: Add MCP servers to .mcp.json and map them to ~~category placeholders in commands. WorkIQ is always on by default.
This project was inspired by SEOMachine by Craig Hewitt. The product-management methodology comes from the Anthropic product-management plugin.
MIT