Skip to content

setup.md

Latest commit

 

History

History
215 lines (146 loc) · 9.83 KB

setup.md

File metadata and controls

215 lines (146 loc) · 9.83 KB
Raw

🚀 PM Brain Setup Guide

Welcome! This guide gets you from zero to running in under 30 minutes.

⚠️ Fork it — but make it private

PM Brain is designed to be forked and made your own — but the original repo is public, so any fork defaults to public too. Your company context, roadmaps, and stakeholder info should NOT be visible to everyone.

Fork it, make it private, then add your context. Don't accidentally push company strategy to the internet.

See Step 1: Get the Repository for exact instructions.

🔧 First: Choose Your AI Tool

Your tool choice affects how PM Brain configures itself — some tools auto-load rules, others require a manual setup prompt each session.

Tool Rules auto-load? Setup effort
Cursor Yes (.cursor/rules/) Minimal
VS Code + GitHub Copilot Yes (.github/copilot-instructions.md) Minimal
Claude Code No — manual per session Some
ChatGPT / Claude.ai No — copy-paste per session More

See platform-setup.md for full per-tool instructions.

The fastest path: once you have the repo open in your AI tool, say "help me set up PM Brain" — the agent reads the repo, flags placeholders, and guides you through the rest.

📋 Quick Start

  1. Get the repo — fork it privately, clone locally, or stay local (Step 1)
  2. Open in your AI tool — IDE (Cursor, VS Code) or web-based (Claude.ai, ChatGPT) — see the AI Tool section above and platform-setup.md for per-platform details
  3. Say "help me set up PM Brain" — the agent reads the repo, flags placeholders, guides you

🎯 Step 1: Get the Repository

Recommended: Fork → make private → clone

  1. Fork this repo on GitHub (top right: Fork)
  2. Make it private: go to your fork → Settings → Danger Zone → Change visibility → Private
  3. Clone your private fork locally: 
    git clone https://github.com/YOUR-USERNAME/pm-brain.git
cd pm-brain
  4. Add upstream to pull future PM Brain updates: 
    git remote add upstream https://github.com/andreaskelm/pm-brain.git

That's it. Your private copy is ready. Changes you push stay in your private repo; you can pull framework updates from upstream whenever you want.

Alternatives

  • Local only (no GitHub): Clone directly — no GitHub account, no privacy risk, no syncing. 
    git clone https://github.com/andreaskelm/pm-brain.git
cd pm-brain
  • Private from the start (no public footprint): Create an empty private repo on GitHub first, then clone PM Brain into it: 
    git clone https://github.com/andreaskelm/pm-brain.git
cd pm-brain
git remote set-url origin https://github.com/YOUR-USERNAME/pm-brain-private.git
git push -u origin main
git remote add upstream https://github.com/andreaskelm/pm-brain.git
  • Team/org repo: Fork into a private org repo instead of a personal one, then follow Step 3 → Team mode.

📚 Step 2: Understand the Structure

Before customizing, understand what each folder does:

pm-brain/
├── 00-Meta/                   # 🧠 Personal practice (daily log, learning log, growth portfolio, Product Judgment Test)
├── 01-Company-Context/        # 🏢 YOUR company vision, strategy, stakeholders
├── 02-Methods-and-Tools/      # 🧭 Frameworks & templates (Strategy, Discovery, Execution, Communication)
├── 03-Research-Artifacts/     # 🔍 Your research notes and findings
└── 04-Initiatives/            # 🚧 Your active product work

Key insight:

  • 00-Meta/ = YOUR PRACTICE (daily log, weekly/monthly reflection, growth portfolio, forecast calibration). Canonical prompts and templates live in 02-Methods-and-Tools/2.0-Foundations/2.0.1-Mental-Models/6-Product-Sense-Development/.
  • 01-Company-Context/ = YOU MUST CUSTOMIZE THIS (your company info)
  • 02-Methods-and-Tools/ = READY TO USE — 2.0-Foundations (think first, product sense), 2.1-Strategy, 2.2-Discovery, 2.3-Execution, 2.4-Communication. Start product thinking from 0-start-here-product-thinking.md.
  • 03-Research-Artifacts/ = YOUR RESEARCH (add your interview notes, findings)
  • 04-Initiatives/ = YOUR WORK (document your active initiatives)

⚙️ Step 3: Choose Public / Private / Team Mode

This repo supports three usage modes. Setup is manual and simple: you choose a mode by appending the right .gitignore template and committing. No scripts required.

Modes at a Glance

  • 🌍 Public (default)

    • Best for: open learning, public examples, blog/portfolio content.
    • Tracked: everything (frameworks, Meta logs, company context, initiatives).

  • 🔒 Private

    • Best for: real company work or personal growth where content must not leak.
    • Tracked: frameworks and docs.
    • Ignored (via .gitignore.private): Meta logs, growth portfolio, company context, active initiatives (per patterns).

  • 👥 Team

    • Best for: a shared frameworks repo, with each person keeping their own practice private.
    • Tracked: 02-Methods-and-Tools/ and any shared docs.
    • Ignored (via .gitignore.team): each contributor’s 00-Meta logs and personal growth artefacts.

How to Apply a Mode (Manual)

Run one of these in your fork’s root:

  • Public (do nothing)

    • Leave .gitignore as‑is. Everything can be tracked.

  • Private mode:

    cat .gitignore.private >> .gitignore
git add .gitignore
git commit -m "setup: configure private mode"

  • Team mode:

    cat .gitignore.team >> .gitignore
git add .gitignore
git commit -m "setup: configure team mode"

For details and decision help, see 00-Meta/MODE-SELECTION-GUIDE.md.

📚 Step 4: Customize

Minimum to start (10 minutes):

  • 01-Company-Context/CONTEXT.md — your name, role, company, team, and 2 sentences on what you build
  • .cursor/rules/thinking.personal.mdc (Cursor / VS Code) — how you work, what you struggle with, communication preferences. For other tools, paste this content as part of your system prompt — see platform-setup.md.

That's it. The agent uses these two files on every conversation.

Add as you go:

  • Full 01-Company-Context/ docs (vision, strategy, stakeholders) when you want grounded strategic advice
  • 01-Company-Context/CONTEXT-HEALTH.md to track freshness once you have multiple docs
  • 01-Company-Context/1.1-Stakeholder-Avatars/ if you want the agent to simulate stakeholder reactions
  • 00-Meta/ daily log + forecast calibration if you want to build product sense deliberately

Not sure what context you have? Start with your team. Fill in what you know, mark gaps with "TBD" or "unclear", and treat the files as working hypotheses — not finished documents. As you learn more, you update them.

If your org has layers (company → division → business unit → team), don't try to model all of them upfront. Start at the level you actually work at — usually your team or business unit — and add layers only when the distinction matters for your work.

A note on git tracking: In private/team modes, CONTEXT.md and thinking.personal.mdc are ignored by git — your personal context stays private. See Step 3 to configure your mode.

🧪 Step 5: Start Using It

Bring a real problem you're working on — a feature you're scoping, a decision you're stuck on, a stakeholder situation you're navigating. Open a chat with your AI tool and describe it.

Don't try to pick the right framework first. The agent's job is to help you think through the mess before suggesting any structure. Lead with the problem; let the conversation go where it needs to go.

The frameworks are there when you need them. 0-start-here-product-thinking.md is the entry point — it has a simple prompt to start, then maps frameworks by situation. Or browse 02-Methods-and-Tools/README.md directly. But the braindump comes first.

📝 Step 6: Set Up Your First Initiative

When you have something worth tracking, create an initiative folder:

  1. Navigate to: 04-Initiatives/

  2. Create a new folder: 4.2-Your-Initiative-Name/ (or use the example: 4.1-Initiative-Codename/)

  3. Copy the structure from the example:

    • README.md
    • summary.md
    • opportunity-assessment.md
    • prd.md
    • roadmap.md
    • decisions.md

  4. Start with: summary.md - Write a brief summary of what you're working on

This is where you'll do your actual product work!

✅ Step 7: You're Ready

Open a chat with your AI tool and say: "verify my PM Brain setup is complete and aligned with best practices" — it will read your context files, check that placeholders are filled, and flag anything missing.

From here, everything is driven by real work and real problems.

📚 Reference