25 product management workflows built from 340 episodes of Lenny’s Podcast.
Bring a pricing page, PRD, interview notes, survey export, or roadmap. Get a usable deliverable: a pricing plan, shaped spec, research synthesis, experiment brief, or decision memo. Each skill diagnoses your situation, picks a framework, works through it, and can grade what you already have.
Built by Udi Menkes · 𝕏 · GenAI PM
All 25 skills · Quickstart · Card gallery · Upgrading from v1 · Contributing
| Bring this | Use this | Ask for this |
|---|---|---|
| A pricing page or plan matrix | price | Rework our packaging and design a willingness-to-pay test. |
| A PRD or rough feature idea | spec | Turn this into a shaped spec, then grade it. |
| Interview notes | customer-interviews | Identify customer forces and opportunities, with evidence. |
| An AI feature and sample outputs | eval-plan | Build an eval plan from the errors in these outputs. |
| A roadmap with too many priorities | prioritize | Score confidence, surface tradeoffs, and propose a sequence. |
| A difficult decision | decide | Write a decision memo with a pre-mortem and next actions. |
The assistant reads what you provide, asks up to four questions for missing context, and works toward the deliverable. Give it your stage, customer, goal, and constraints when you can.
Run these commands inside Claude Code:
/plugin marketplace add menkesu/awesome-pm-skills
/plugin install awesome-pm-skills@awesome-pm-skills
Then invoke a skill by its plugin name:
/awesome-pm-skills:price
Here’s our pricing page. Grade it and propose three improvements.
The plugin contains the 25 skills under skills/. It needs no API key, external service, or additional runtime to read its instructions. Your assistant supplies the model and any tools it uses to work with your files. See Claude Code’s plugin documentation.
Clone the repository, then copy one skill into your personal skills directory:
git clone https://github.com/menkesu/awesome-pm-skills.git
mkdir -p ~/.agents/skills
cp -R awesome-pm-skills/skills/price ~/.agents/skills/To install all 25, replace the final command with:
cp -R awesome-pm-skills/skills/* ~/.agents/skills/Check for existing folders with the same names before copying; copying can overwrite files. For team use, copy into .agents/skills/ in your project instead. Codex and Cursor discover skills there; see the official Codex and Cursor guides. In Codex, invoke $price; in Cursor, use /price or describe the task so the assistant can select the skill.
After cloning, use ~/.claude/skills/ for personal skills or .claude/skills/ for project skills:
mkdir -p ~/.claude/skills
cp -R awesome-pm-skills/skills/price ~/.claude/skills/Invoke /price. Choose either the plugin or the copy method to avoid duplicate installations.
Open the relevant SKILL.md and provide it with your work. Include that skill’s references/frameworks.md and references/quotes.md when you want the full supporting material. Automatic discovery depends on the assistant; simply cloning this repository does not install the skills into every product.
More examples and local plugin testing →
- Context first. Stage, customer, goal, and constraints route you to an appropriate framework.
- A workflow to run. Concrete steps, research questions, thresholds, and a fill-in deliverable.
- Named disagreements. Where guests differ, the skill explains the conditions that favor each approach.
- A review mode. Grade an existing artifact and get the three most useful fixes.
- Source material. Attributed frameworks and transcript-matched quotes with timestamps.
- An illustrated lead team. The card travels with the skill in its
assets/folder.
Each skill has three lead guests and draws on other guests who discussed the topic. The portraits are AI-generated illustrations based on reference photos; they do not imply endorsement.
| Skill | You get | Lead team |
|---|---|---|
eval-plan |
Eval suite for an AI feature: error analysis → criteria → LLM judges | Hamel Husain · Shreya Shankar · Sander Schulhoff |
agent-workflow |
Your agent setup: lanes, plan mode, autonomy ladder | Claire Vo · Boris Cherny · Aishwarya Naresh Reganti |
ai-prototype |
Idea → working prototype, with build prompts and a debugging loop | Lazar Jovanovic · Zevi Arnovitz · Guillermo Rauch |
ai-product-bets |
Which AI features to bet on, and whether they get better as models do | Nick Turley · Kevin Weil · Dianne Penn |
| Skill | You get | Lead team |
|---|---|---|
strategy |
Strategy one-pager + moat stress test | Roger Martin · Richard Rumelt · Hamilton Helmer |
prioritize |
Confidence-scored roadmap + goals | Itamar Gilad · Christina Wodtke · Janna Bastow |
decide |
Decision memo: eigenquestion, pre-mortem, reversibility | Annie Duke · Shreyas Doshi · Shishir Mehrotra |
spec |
Shaped spec or PR/FAQ, graded | Ryan Singer · Bill Carr · Lane Shackleton |
| Skill | You get | Lead team |
|---|---|---|
customer-interviews |
Interview guide → synthesis of forces, jobs and opportunities | Teresa Torres · Bob Moesta · Judd Antin |
validate-idea |
Riskiest assumptions + the cheapest test for each | Mike Maples Jr. · Eric Ries · Jake Knapp |
pmf-check |
Product-market-fit diagnosis from your survey or usage data | Sean Ellis · Todd Jackson · Rahul Vohra |
| Skill | You get | Lead team |
|---|---|---|
price |
Pricing model, tiers, willingness-to-pay plan | Madhavan Ramanujam · Naomi Ionita · Patrick Campbell |
position |
Positioning canvas + strategic narrative | April Dunford · Andy Raskin · Arielle Jackson |
growth-model |
Growth loops, activation and retention plan | Elena Verna · Casey Winters · Lauryn Isford |
experiment |
Experiment brief + trustworthiness checks + readout | Ronny Kohavi · Ramesh Johari · Archie Abrams |
metrics |
North star + metric tree | Sarah Tavel · Crystal Widjaja · Jessica Lachs |
launch |
Launch plan, story and distribution | Emilie Gerber · Lulu Cheng Meservey · Jason Feifer |
b2b-sales |
Founder-led sales motion: ICP, pitch, pipeline | Jen Abel · Geoffrey Moore · Pete Kazanjy |
| Skill | You get | Lead team |
|---|---|---|
exec-comms |
Exec update, memo or talk, rewritten | Wes Kao · Nancy Duarte · Matt Abrahams |
hard-conversations |
Feedback / hard-conversation script | Kim Scott · Carole Robin · Alisa Cohn |
influence |
Stakeholder map + influence plan | Jessica Fain · Jeffrey Pfeffer · Hilary Gridley |
hire |
Role scorecard + interview loop + references | Adam Ward · Keith Rabois · Lauren Ipsen |
career |
Career plan, promotion case, job-move decision | Phyl Terry · Nikhyl Singhal · Deb Liu |
craft-review |
Quality review: friction log, critical journeys | Katie Dill · Dylan Field · Stewart Butterfield |
ship-faster |
Diagnosis of what slows delivery → velocity plan | Farhan Thawar · Nicole Forsgren · Melissa Perri |
The collection was rebuilt around 25 PM jobs. Every skill now includes diagnosis, framework routing, execution steps, expert disagreements, a deliverable template, and a grading rubric. The canonical folders moved to skills/<name>/, with supporting references and guest cards in each folder.
If you installed v1, use the migration guide to choose replacements and remove obsolete copies from your assistant’s skills directory. The original v1 content remains available in Git history. The separate One Step Better AI PM skill is retained as an optional extra; it requires a GenAI PM subscription and is not included in the 25-skill plugin.
The build used a 340-episode corpus and 10,567 extracted insights. The published skills include 978 checked quoted passages, with guest attribution and timestamps. The quote checker compares quoted passages against a local transcript corpus after normalizing whitespace, case, and common punctuation. Explicit ellipses are checked fragment by fragment. Matching text does not independently verify attribution, timestamps, or the interpretation of a framework.
Transcripts are not bundled. To repeat the check with your own copy:
python3 check_quotes.py skills --transcripts /path/to/transcriptsValidation and provenance details →
Report a misleading claim, broken link, missing attribution, or confusing workflow with the skill name and a concrete example. For edits, follow the contributor guide and run the repository validator:
python3 -m pip install -r requirements-dev.txt
python3 scripts/validate_repo.pyFrameworks, ideas, and quoted language are credited to their original creators throughout the skills. Thank you to Lenny Rachitsky, his guests, and ChatPRD’s transcript project.
This is an independent community project, unaffiliated with and not endorsed by Lenny’s Podcast or the featured guests. The repository’s authored material is covered by the MIT license; attributed third-party material retains its original ownership.
Made by Udi Menkes · GenAI PM