Claude Code Setup for Product Managers

    Claude Code configuration for product managers. Covers PRD writing, sprint planning, stakeholder communication, user research synthesis, and metrics tracking.

    intermediate
    82 views
    productsaasroadmapsprintsuser-research

    CLAUDE.md

    CLAUDE.mdpaste in your project root
    # CLAUDE.md — Product Manager
    > Version: 2026-02-17 | Difficulty: intermediate
    
    ## What Your AI Can Now Do
    - Draft a full PRD from a problem statement — complete with bet, success metric, scope, and non-goals
    - Break any PRD into engineering tickets with acceptance criteria and story points
    - Generate sprint status reports by pulling data from GitHub issues and PRs
    - Pre-load customer context before every call so you walk in prepared
    - Flag scope creep early by comparing sprint commitments against actual progress
    
    ---
    
    ## Quick Start
    1. Define your team and product area in the About Me section below
    2. Run `/prd` with a problem statement to see a full PRD draft in seconds
    3. Connect GitHub and Google Drive in `.env` for live data pulls
    4. Add your current roadmap to `product/roadmap/` so Claude knows your priorities
    5. Run `/status` on Monday morning to get your first automated sprint report
    
    ---
    
    ## About Me
    B2B SaaS product manager owning roadmap, sprint planning, and activation/retention metrics.
    Team: 2 designers, 8 engineers, 1 data analyst. We ship bi-weekly.
    I think in bets, not features — every initiative ties to a measurable outcome.
    
    ---
    
    ## Pre-Built Commands
    
    | Command | What It Does |
    |---------|-------------|
    | `/prd` | Draft a PRD from a problem statement — bet, metrics, scope, non-goals, and open questions |
    | `/tickets` | Break a PRD into engineering tickets with acceptance criteria and story points |
    | `/status` | Generate sprint or release status from GitHub issues and merged PRs |
    | `/customer-prep` | Pull account context and recent support tickets before a customer call |
    | `/release-notes` | Draft user-facing release notes from shipped PRs and closed issues |
    
    ---
    
    ## How Claude Connects to Your Tools
    
    | Tool | Why You Need It |
    |------|----------------|
    | GitHub | PR and issue tracking — the source of truth for what actually shipped |
    | Google Workspace | PRDs, research docs, stakeholder updates, competitive analysis |
    | Google Calendar | Sprint ceremonies, customer calls, design reviews |
    | Memory | Customer patterns, team velocity history, stakeholder preferences |
    
    ### Setup
    ```bash
    # Required in .env
    GITHUB_TOKEN=your-token
    GOOGLE_WORKSPACE_EMAIL=your-email@company.com
    ```
    
    ---
    
    ## Product Rules
    - **MUST:** Every PRD starts with a one-line "bet we're making" before any feature detail
    - **MUST:** Tag every feature with the metric it moves — no metric, no prioritization
    - **MUST:** Talk to at least 3 customers before writing any PRD — internal intuition is not validation
    - **SHOULD:** Keep tickets under 300 words — if it needs more, the scope is wrong
    - **MUST:** Answer engineer questions in PR reviews within 2 hours, not "next sync"
    - **SHOULD:** Write the problem and acceptance criteria in tickets — let engineers design the solution
    - **MUST:** Customer quotes go in verbatim — never paraphrase pain
    - **SHOULD:** Lead stakeholder emails with the decision needed, not the background
    
    ---
    
    ## Project Structure
    ```
    product/
      roadmap/
      prds/
      research/
      competitive/
      releases/
      metrics/
    ```
    
    ### What These Folders Mean
    - `roadmap/` — Quarterly roadmap and prioritization frameworks tied to business goals
    - `prds/` — All PRDs organized by quarter (e.g., 2026-Q1-feature-name.md)
    - `research/` — Customer interviews, survey results, and analytics deep dives
    - `competitive/` — Competitor teardowns and market positioning analysis
    - `releases/` — Release notes drafts and changelog history
    - `metrics/` — Dashboard definitions, experiment results, and metric baselines
    
    ---
    
    ## Automation
    
    | When | What Happens | If It Fails |
    |------|-------------|-------------|
    | PRD saved to Drive | Auto-extract open questions and create a checklist issue in GitHub | Manually review PRD for unresolved questions and create issue |
    | Sprint ends (bi-weekly) | Pull shipped PRs and closed issues into draft release notes | Export PR list from GitHub and draft notes manually |
    | Customer call on calendar | Pre-load account context and recent support tickets | Check CRM and support tool manually before the call |
    | GitHub PR merged | Update relevant PRD status from "In Development" to "Shipped" | Manually update PRD status during weekly review |
    | Monday morning | Generate sprint status with at-risk items highlighted | Review GitHub project board and flag blockers in standup |
    
    ---
    
    ## Gotchas
    - **MUST:** "Fast follow" is a lie — if it matters, put it in v1. Fast follows ship 30% of the time
    - **MUST:** Metrics without baselines are decoration — measure current state before launching any "improvement"
    - **SHOULD:** Bi-weekly shipping means cutting scope, not cramming — if it's Wednesday and scope is at risk, cut
    - **MUST:** Never confuse shipping with impact — track the metric after launch, not just the deploy
    - **SHOULD:** Roadmap is a communication tool, not a commitment — update it when reality changes
    
    ---
    
    ## Memory & Learning
    Claude builds context over time:
    - Product domain knowledge and team velocity patterns across sprints
    - Recurring stakeholder preferences and communication styles
    - Customer pain points and feature request patterns from logged calls
    - What slipped in past sprints and why — used to flag scope creep early
    - PRD structure preferences and which formats get engineering buy-in
    
    To save something important: just say "remember that [fact]" in conversation.
    To review saved memories: use the `/memory` command to open the memory editor.
    Each setup follows the 5-layer Context Engineering framework: Identity, Memory, MCP, Skills, and Hooks.
    Learn the framework

    Pre-Built Skills

    Ready-to-use skills included in this setup. Each runs a specialized workflow for your role.

    📋/prd

    Draft product requirements from user research

    🎯/prioritize

    Stack rank features by impact and effort

    📝/release-notes

    Write user-facing release notes from changelog

    🔍/user-research

    Synthesize interview notes into insights

    Recommended MCP Servers

    Gmail MCP

    Stakeholder updates and customer feedback

    Calendar MCP

    Sprint ceremonies and user interview scheduling

    GitHub MCP

    Track issues, PRs, and engineering velocity

    Google Drive MCP

    Shared PRDs, research docs, and roadmap

    Slack MCP

    Team communication and async decisions

    Workflow Templates

    Write PRD

    /prd

    Generate a product requirement document from a feature brief

    Sprint Tickets

    /tickets

    Break a PRD into engineering tickets with acceptance criteria

    Stakeholder Update

    /status

    Generate weekly status update for leadership

    This Is a Starter Config

    MemoryOS analyzes YOUR actual workflow, tools, and communication style to generate a brain personalized to you. Fully automated — you decide what to share, it does the rest.

    Based on patterns from 100+ professionals. No technical knowledge required.

    Get Your Personalized Setup