Open source · For product managers
Claude Code configured for PMs to ship faster.
Templates, skills, and ten practices that turn a fresh Claude Code session into a working PM workflow. Clone the repo, fill in three files, ship.
MIT licensed. Apache-2.0 attribution included in every shipped skill.
10 best practices
What PMs shipping fastest with Claude Code do differently.
One screen per practice. Read in order, or jump to the one that solves the problem you have today.
Write the PRD before the prompt
A PRD answers the four things a chat prompt skips: problem, user, done, and out-of-scope.
Use plan mode for anything non-trivial
Plan mode forces Claude to think before it edits. Approve the plan, then implementation gets boring.
Make CLAUDE.md the single source of truth
Product, voice, architecture, and conventions live in one file Claude reads on every session.
Author skills, not prompts
Skills auto-load by trigger phrase. Stop re-typing the same instructions in chat.
Ship a DESIGN.md with brand tokens
Tokens at the source level so contributors can’t reach default indigo by accident.
Wire up the right MCPs early
GitHub, Linear, Notion, PostHog. The model can read real data, not your paraphrase of it.
Never commit a secret
.env.example in repo, .env in .gitignore, MCP tokens loaded from the shell.
Treat the design system as a skill
Colors, type, voice, surfaces. Claude defaults to indigo and mesh gradients without it.
Generate product animations with Remotion
Programmatic video lets one PM ship feature demos and onboarding walkthroughs faster than waiting for the design team.
Onboard the agent like a teammate
First-day reading list, code map, voice doc. Treat the agent like a new hire who reads fast.
Log every session so Claude can pick up where you left off
Two files per task — requirements and a running work log — so Claude starts each session with full context instead of a blank slate.
Templates
The files Claude reads on every session.
Three of the ten in the repo. The whole set lives on /templates.
templates/CLAUDE.mdView on GitHub# CLAUDE.md
> Memory file. Claude reads this on every session start. Keep it under 500 lines.
> Replace the bracketed TODOs with details about your team and product.
## Product
[TODO] One paragraph: what you're building and who it's for.
Example: *We build a payments API for marketplaces with 1–50 sellers. Our buyer is the marketplace's founding engineer. Our anti-buyer is enterprise procurement.*
## Team
- **PM:** [TODO — your name + handle]
- **Eng lead:** [TODO]
- **Design:** [TODO]
- **Anyone else with commit access:** [TODO]
…templates/DESIGN.mdView on GitHub# DESIGN.md
> Brand spec template. Fill in your brand values and Claude reads them every session.
> Reference this file from `CLAUDE.md` so the design system is always in context.
>
> For the deeper version (used by Claude when generating artifacts), see `skills/your-brand/SKILL.md`.
## At a glance
| Element | Spec |
|---|---|
| Primary color | [TODO `#xxxxxx`] |
| Logo | [TODO file path + dark/light variants] |
| Headline type | [TODO size/weight/letter-spacing/color] |
| Body type | [TODO size/line-height/color] |
| Voice | [TODO one sentence — e.g. "Short, declarative, period-ending. One idea per sentence."] |
## Colors
…pr_flow/create-prd.mdcView on GitHub---
description: PRD authoring rule. Invoke when the user wants a new feature spec.
globs:
alwaysApply: false
---
# Rule: Generating a Product Requirements Document (PRD)
## Goal
To guide an AI assistant in creating a detailed Product Requirements Document (PRD) in Markdown format, based on an initial user prompt. The PRD should be clear, actionable, and suitable for a junior developer to understand and implement the feature.
## Process
1. **Receive Initial Prompt:** The user provides a brief description or request for a new feature or functionality.
2. **Ask Clarifying Questions:** Before writing the PRD, the AI *must* ask clarifying questions to gather sufficient detail. The goal is to understand the "what" and "why" of the feature, not necessarily the "how" (which the developer will figure out). Make sure to provide options in letter/number lists so the user can respond easily with selections.
3. **Generate PRD:** Based on the initial prompt and the user's answers, generate a PRD using the structure outlined below.
4. **Save PRD:** Save the generated document as `tasks/[feature-name]/task.md`.
5. **Seed sessions log:** Create `tasks/[feature-name]/sessions.md` with a single starter entry: today's date, status "Not started", and any open questions noted.
…Who says so
Synthesized from PMs and engineers shipping with Claude every day.
Aakash Gupta
Product Growth · Albuquerque podcast
“The PMs winning right now treat Claude Code like a junior dev, not a search bar.”Listen →
Aman Khan
PM at Arize · Every PM should be building skills
“Skills encode judgment. A prompt is a one-time ask; a skill is a teammate that shows up.”Read →
Cat Wu
Claude Code PM at Anthropic
“Our team writes a spec before the code. Then the code writes itself, mostly.”Read on Lenny’s →
Ryan Nystrom
EM at Notion · via chatprd.ai
“A clean voice doc plus plan mode equals a PR in twenty minutes, not a sprint.”Read →
Owen Williams
Design Engineer at Stripe · How I AI
“The design system has to be a skill. Otherwise every page comes back indigo.”Watch →
Get started
Clone the repo. Open it in Claude Code.
Two minutes from clone to your first PRD. Fill in three files and ship.
git clone https://github.com/runtm-ai/claudecode-for-pms.gitMIT licensed. Star the repo if it saves you an afternoon.