Status: stable. 121 skills authored Feb–Mar 2026.
Your AI agent can write code, summarize documents, and answer questions. But ask it to size an experiment, diagnose a metric drop, or build a launch plan — and it wings it. Generic answers, hallucinated numbers, skipped steps.
This library fixes that.
121 structured skills across 15 product domains. Each skill is a playbook that tells the agent exactly what to ask, how to reason through the problem, and what the output should look like. No prompt engineering on your side — just install a domain and go.
Think of it as product management muscle memory for your AI agent. The same frameworks, the same rigor, the same questions a great PM would ask — packaged as reusable skills.
- Open Customize (bottom-left)
- Go to Browse plugins → Personal → +
- Select Add marketplace from GitHub
- Enter:
alexe-ev/product-plugins
All 15 plugins install automatically, providing both commands and skills.
# Step 1: Add the marketplace
/plugin marketplace add alexe-ev/product-plugins
# Step 2: Install individual plugins
/plugin install experimentation@product-plugins
/plugin install data-analytics@product-plugins
/plugin install product-strategy@product-plugins
/plugin install product-discovery@product-plugins
/plugin install product-planning@product-plugins
/plugin install gtm@product-plugins
/plugin install marketing-growth@product-plugins
/plugin install ai-product@product-plugins
/plugin install business-commercial@product-plugins
/plugin install customer-success@product-plugins
/plugin install product-ops@product-plugins
/plugin install leadership-communication@product-plugins
/plugin install technical-product@product-plugins
/plugin install risk-compliance@product-plugins
/plugin install ux-research@product-pluginsThe skills/*/SKILL.md files follow the universal skill format. Copy skill folders to the appropriate tool directory:
| Tool | Directory | Command |
|---|---|---|
| Gemini CLI | ~/.gemini/skills/ |
for plugin in */; do cp -r "$plugin/skills/"* ~/.gemini/skills/ 2>/dev/null; done |
| OpenCode | .opencode/skills/ |
for plugin in */; do mkdir -p .opencode/skills/; cp -r "$plugin/skills/"* .opencode/skills/ 2>/dev/null; done |
| Cursor | .cursor/skills/ |
Copy skill folders manually |
| Codex CLI | .codex/skills/ |
Copy skill folders manually |
| Kiro | .kiro/skills/ |
Copy skill folders manually |
/experimentation:estimate-sample-size
The skill walks Claude through the right questions, methodology, and output format. If required inputs are missing, it stops and asks — no hallucinated numbers, no guesswork.
Running an A/B test and need to know if you have enough traffic:
/experimentation:estimate-sample-size
Metric dropped 12% last week and nobody knows why:
/data-analytics:diagnose-metric-movement
Investor deck needs a real business case, not vibes:
/business-commercial:build-business-case
Stakeholders are misaligned on the roadmap, again:
/product-planning:build-roadmap-prioritization
No googling, no guessing, no "let me think about that." Each skill walks Claude through the right questions, the right methodology, and the right output format — every time.
| Domain | What it covers | Skills | |
|---|---|---|---|
| 🧪 | experimentation |
Hypothesis design, sample sizing, experiment plans, result analysis, post-test decisions | 8 |
| 📊 | data-analytics |
Metrics frameworks, funnel analysis, dashboards, signal detection, metric diagnosis | 8 |
| 🧭 | product-strategy |
Vision, long-term direction, market fit, positioning, competitive strategy, portfolio | 9 |
| 🔍 | product-discovery |
Problem framing, user interviews, JTBD, insight synthesis, opportunity identification | 8 |
| 📋 | product-planning |
OKRs, roadmapping, prioritization, PRDs, user stories, delivery planning | 8 |
| 🚀 | gtm |
Launch planning, positioning, messaging, pricing, sales battlecards, enablement | 8 |
| 📈 | marketing-growth |
Growth models, acquisition funnels, lifecycle, onboarding, conversion, channel experiments | 8 |
| 🤖 | ai-product |
AI feature ideation, model tradeoffs, HITL workflows, quality monitoring, value framing | 8 |
| 💰 | business-commercial |
Business cases, unit economics, LTV/CAC, forecasting, ROI, revenue strategy | 8 |
| 🤝 | customer-success |
Churn analysis, health scoring, VOC programs, feedback triage, retention playbooks | 8 |
| ⚙️ | product-ops |
Operating cadence, planning processes, tooling, cross-team alignment, wikis | 8 |
| 🎤 | leadership-communication |
Executive narratives, stakeholder mapping, decision frameworks, team rituals | 8 |
| 🔧 | technical-product |
Platform thinking, architecture-aware prioritization, engineering collaboration, NFRs | 8 |
| 🛡️ | risk-compliance |
Risk registers, privacy assessment, ethical frameworks, responsible AI governance | 8 |
| 🎨 | ux-research |
Research planning, usability testing, journey mapping, experience quality, design handoff | 8 |
Install any combination:
/plugin marketplace add alexe-ev/product-plugins
/plugin install experimentation@product-plugins
/plugin install data-analytics@product-plugins
/plugin install product-strategy@product-pluginsFull skill list by domain
| Skill | Description |
|---|---|
formulate-experiment-hypothesis |
Turn a product idea into a testable hypothesis |
validate-hypothesis-quality |
Check if a hypothesis is specific, measurable, and testable |
define-success-metrics |
Define primary metric, secondary metrics, and guardrails |
estimate-sample-size |
Calculate required sample size and experiment duration |
design-experiment-plan |
Build a complete experiment plan from hypothesis to execution |
analyze-experiment-results |
Analyze control vs test results with statistical rigor |
validate-experiment-quality |
Audit whether an experiment was run correctly |
recommend-post-test-decision |
Recommend next steps based on experiment outcome |
| Skill | Description |
|---|---|
design-product-metrics |
Define a metrics framework with primary, secondary, and guardrail metrics |
analyze-funnel-retention-cohorts |
Analyze funnel drop-off, retention curves, and cohort behavior |
build-decision-dashboard |
Structure a dashboard that supports fast decision-making |
design-metric-alert-system |
Design alerting with thresholds, logic, and escalation paths |
detect-performance-signals |
Identify and interpret signals of product performance change |
diagnose-metric-movement |
Diagnose root cause of unexpected metric movements |
plan-event-instrumentation |
Plan event tracking and instrumentation for product analytics |
run-cohort-analysis |
Structure and interpret cohort analysis for retention and behavior |
| Skill | Description |
|---|---|
define-product-vision |
Articulate a clear product vision |
craft-product-mission |
Define the product mission statement |
define-long-term-direction |
Translate vision into strategic priorities and trade-offs |
define-strategic-positioning |
Define how the product is positioned in the market |
assess-product-market-fit |
Evaluate product-market fit signals |
analyze-competitive-landscape |
Map competitive landscape and dynamics |
frame-competitive-differentiation |
Frame what differentiates the product from alternatives |
build-portfolio-roadmap-strategy |
Build a portfolio-level roadmap strategy |
prioritize-strategic-bets |
Prioritize strategic bets and investment areas |
| Skill | Description |
|---|---|
identify-problem-opportunity |
Frame the real problem or opportunity before jumping to solutions |
scope-discovery-research |
Define scope, goals, and methods for discovery research |
run-user-interviews |
Plan, structure, and synthesize user interviews |
apply-jtbd-framework |
Apply Jobs-to-be-Done to understand what job users hire the product for |
synthesize-qualitative-research |
Synthesize interviews and qualitative data into actionable insights |
generate-insights |
Distill raw observations into meaningful product insights |
frame-insight-opportunity |
Turn insights into structured opportunity statements |
develop-persona-segment |
Build and validate user personas and segments |
| Skill | Description |
|---|---|
set-goals-okrs-kpis |
Define and structure goals using OKRs and KPIs |
align-goals-to-strategy |
Check alignment between team goals and company strategy |
build-roadmap-prioritization |
Build and prioritize a feature or initiative roadmap |
apply-prioritization-framework |
Apply RICE, MoSCoW, ICE, or opportunity scoring |
write-requirements-prd |
Write complete product requirements documents |
write-user-stories |
Write user stories with clear acceptance criteria |
plan-delivery-collaboration |
Structure delivery planning and cross-functional alignment |
run-retrospective |
Facilitate a retrospective to extract learning |
🚀 gtm
| Skill | Description |
|---|---|
plan-product-launch |
Build a launch plan with timeline, channels, and coordination |
build-launch-readiness-checklist |
Cover all cross-functional dependencies before launch |
develop-positioning-messaging |
Develop positioning and messaging for target segments |
craft-value-proposition |
Craft a clear value proposition for a product or feature |
design-packaging-tiers |
Design packaging tiers aligned with value and willingness to pay |
support-pricing-packaging |
Analyze and recommend pricing structures |
create-sales-battlecard |
Create competitive battlecards for sales |
enable-sales-cs |
Structure sales and CS enablement materials |
| Skill | Description |
|---|---|
design-growth-model |
Build or assess a growth model (AARRR) |
map-acquisition-funnel |
Map acquisition funnel from awareness to activation |
identify-growth-loop |
Identify and map compounding growth loops |
optimize-conversion-growth-loops |
Identify conversion bottlenecks and reinforce growth loops |
design-onboarding-flow |
Design onboarding that gets users to activation fast |
plan-lifecycle-engagement |
Design lifecycle engagement including onboarding and re-engagement |
design-channel-experiment |
Design a structured experiment for a new growth channel |
strategize-channel-campaigns |
Develop channel strategy and campaign experimentation plan |
| Skill | Description |
|---|---|
ideate-ai-features |
Generate AI feature ideas with value propositions and feasibility |
assess-model-capabilities |
Evaluate AI model capabilities and limitations for a product decision |
make-model-tradeoff-decision |
Structure trade-offs between AI models or approaches |
frame-ai-product-value |
Frame AI feature value in terms of user outcomes |
design-human-in-loop-workflow |
Balance AI automation with human oversight |
define-ai-success-metrics |
Define success metrics covering quality, UX, business, and safety |
evaluate-ai-quality-monitoring |
Set up AI quality monitoring and evaluation loops |
run-ai-prototype-evaluation |
Evaluate an AI prototype for quality and readiness |
| Skill | Description |
|---|---|
build-business-case |
Develop a structured business case with ROI and risks |
frame-roi-analysis |
Frame ROI analysis for a product initiative |
analyze-unit-economics |
Analyze CAC, LTV, payback period, and gross margin |
model-ltv-cac |
Model LTV and CAC for unit economics assessment |
evaluate-pricing-model |
Evaluate pricing model against behavior and goals |
align-revenue-strategy |
Align product decisions with revenue model and commercial goals |
run-forecasting-scenarios |
Build and compare forecasting scenarios |
run-sensitivity-analysis |
Identify which assumptions have the most impact |
| Skill | Description |
|---|---|
analyze-churn-retention |
Analyze churn patterns and design retention interventions |
build-customer-health-score |
Build a health score model predicting churn and expansion |
monitor-adoption-health |
Define and monitor adoption health metrics |
run-voc-program |
Design a Voice of Customer program for systematic insights |
design-customer-feedback-survey |
Design surveys that generate reliable, actionable data |
synthesize-feedback-themes |
Synthesize feedback volumes into structured themes |
triage-feedback-loop |
Establish feedback intake and triage routing |
design-retention-playbook |
Design retention playbook for at-risk accounts |
⚙️ product-ops
| Skill | Description |
|---|---|
design-operating-cadence |
Design operating rhythm: meetings, reviews, reporting, planning |
design-planning-process |
Connect strategy to execution through repeatable planning |
align-cross-team-communication |
Design communication systems that keep teams aligned |
run-cross-functional-review |
Facilitate cross-functional product review for alignment |
audit-product-process |
Identify bottlenecks and improvement opportunities |
automate-workflow-governance |
Identify automation opportunities and governance standards |
manage-tooling-documentation |
Audit tooling stacks and documentation practices |
build-product-wiki |
Design and populate a product knowledge base |
| Skill | Description |
|---|---|
build-executive-narrative |
Craft executive-level narratives about strategy and results |
write-product-strategy-doc |
Write a product strategy document that aligns teams |
communicate-decisions-tradeoffs |
Communicate decisions and trade-offs to build trust |
apply-decision-framework |
Apply DACI, SPADE, RAPID for clear decisions |
map-stakeholder-influence |
Map stakeholders and develop engagement strategies |
run-stakeholder-alignment |
Execute stakeholder alignment for decisions or initiatives |
build-operating-norms |
Establish team operating norms and working agreements |
design-team-rituals |
Design rituals and rhythms that build alignment |
| Skill | Description |
|---|---|
run-technical-discovery |
Surface feasibility constraints and architectural options |
collaborate-with-engineering |
Structure product-engineering collaboration patterns |
define-non-functional-requirements |
Define performance, security, scalability, accessibility, compliance |
prioritize-architecture-aware |
Incorporate architectural constraints into prioritization |
manage-technical-debt-tradeoffs |
Structure trade-offs between tech debt and new value |
develop-platform-product-thinking |
Define developer experience, extensibility, ecosystem strategy |
design-developer-experience |
Design DX for platform, API, or SDK products |
assess-reliability-scalability |
Evaluate reliability and scalability for product decisions |
| Skill | Description |
|---|---|
run-pre-mortem |
Identify ways an initiative could fail before launch |
design-risk-register |
Design and populate a risk register for an initiative |
plan-risk-mitigation |
Develop mitigation plans with owners, triggers, contingencies |
assess-privacy-security-regulatory |
Identify privacy, security, and regulatory risks |
run-privacy-impact-assessment |
Run a PIA for features that handle personal data |
apply-ethical-decision-framework |
Apply ethical frameworks to evaluate product decisions |
conduct-ai-audit |
Audit AI systems against responsible AI standards |
govern-responsible-ai |
Establish governance for responsible AI development |
| Skill | Description |
|---|---|
plan-ux-research |
Plan UX research initiatives |
design-research-study |
Design a structured research study |
run-usability-testing |
Plan and run usability testing sessions |
analyze-usability-findings |
Analyze and synthesize usability test findings |
map-user-journey |
Map the end-to-end user journey |
assess-experience-quality |
Assess overall experience quality |
manage-design-handoff |
Structure design-to-engineering handoff |
write-design-brief |
Write a clear design brief |
All 121 skills were evaluated across 3 independent rounds. Each round used different test cases and evaluation criteria written by a different LLM without access to skill instructions. Claude evaluated responses in all rounds — which matches real-world usage but means results carry some self-evaluation bias.
Overall results across 3 rounds (poor-context inputs):
| Round 1 | Round 2 | Round 3 | Avg | |
|---|---|---|---|---|
| Avg WITH skill | 95% | 94% | 99% | 96% |
| Avg WITHOUT skill | 84% | 54% | 37% | 58% |
| Avg delta | +11 pp | +40 pp | +62 pp | +38 pp |
The spread across rounds reflects how strictly each LLM wrote the criteria. The average delta of +38 pp is the most reliable single number.
Where skills add the most value:
Skills that enforce structured output, block on missing inputs, or guard against specific failure modes consistently outperform vanilla Claude. Strongest results in risk-compliance, gtm (launch plans, battlecards), data-analytics, and product-ops.
Where skills add less:
Domains where Claude's training already covers the framework well — experimentation hypothesis formulation, general product-strategy — show smaller gains. Vanilla Claude handles these reasonably without additional guidance.
What skills prevent in practice:
- Hallucinated specifics. Without a skill, the model invents numbers, thresholds, and recommendations on vague inputs. Skills stop and ask instead.
- Skipped prerequisites. Without a skill, the model proceeds with incomplete inputs. Skills surface what's missing before producing output.
- Wrong output format. Without a skill, responses are conversational. Skills enforce the structure a PM can actually hand off.
Every skill is a standalone playbook:
experimentation/skills/estimate-sample-size/
├── SKILL.md ← what to ask, how to reason, what to output
├── REFERENCE.md ← formulas and methodology (calculation skills)
└── examples/ ← how the skill behaves with complete vs. incomplete inputs
Skills don't hallucinate methodology. If required inputs are missing, they stop and ask. If the math is uncertain, they say so.
Part of the layer underneath AI products: the tools measure and route (ai-economics, ai-ab-testing-tool, multi-model-router, ml-impact-calculator); this repo packages the PM expertise itself so agents can execute it. Built by the maker of Whisperly.
Contributions are welcome. Whether it's improving an existing skill, adding a new one, or benchmarking another domain — open a PR or issue.
When adding or editing a skill:
- Follow the existing structure:
SKILL.md+ optionalREFERENCE.md+examples/ - Include at least one example with complete inputs and one with missing inputs
- Skills should ask for missing information, not hallucinate it
- Keep the scope tight — one skill, one job