product-showcase
Generate a comprehensive marketing website for a web app — multi-page with real screenshots, animated GIF walkthroughs, feature deep-dives, and workflow demonstrations.
Install / Use
npx skills add jezweb/claude-skills --skill product-showcaseInstalls into whichever agent you are using.
SKILL.md
Installable skill definition
Quality Score
Category
AutomationSupported Platforms
Tags
Our assessment of product-showcase
product-showcase scores 88/100 on our quality scale, 1674th of 2,888 Automation skills we index.
Its SKILL.md is 19 KB long, well organised into 44 sections with 17 code examples: a thorough specification that gives an agent plenty to work with.
With 1,025 GitHub stars, it is one of the more widely adopted skills in the catalogue.
Maintenance, license and trust
- The repository was last updated about 3 months ago. That is recent enough to be usable, but agent tooling moves fast, so check the instructions against your agent's current version.
- Our last check on 2026-10-10 found the source still online.
- It is released under the MIT license, a permissive license that allows use, modification and commercial use with attribution.
- Its trust signals score 98/100, with no cautions. These come from repository metadata, not a code audit — read the skill file before letting an agent act on it.
Safety scan
No issues foundOur scan of the whole file found no instruction hijacking, hidden characters, credential access, data exfiltration or destructive commands. An AI review of the same text found nothing harmful.
AI review by kimi-k2.7-code on 2026-10-06. Automated pattern scan on 2026-10-06. It catches known dangerous patterns, not every risk — read a skill before letting an agent act on it.
product-showcase compared with similar skills
All 4 of these similar skills score higher than product-showcase; compare them before choosing.
| Skill | Score | Stars | Updated | Format |
|---|---|---|---|---|
| product-showcase (this skill)by jezweb | 88 | 1.0k | 3mo ago | SKILL.md |
| Agent-Reachby Panniantong | 100 | 96.8k | 5d ago | CLAUDE.md |
| Scraplingby D4Vinci | 100 | 87.1k | today | MCP Server |
| rufloby ruvnet | 100 | 74.5k | today | MCP Server |
| algorithmic-artby anthropics | 100 | 177.9k | 20d ago | SKILL.md |
Frequently asked questions
- How do I install product-showcase?
- Run
npx skills add jezweb/claude-skills --skill product-showcase. The install tabs above show the steps for each supported agent. - Which AI agents does product-showcase work with?
- It is written for Universal, as a SKILL.md file. Other agents that read the same format can often use it too.
- Is product-showcase safe to use?
- Our scan of the whole file found no instruction hijacking, hidden characters, credential access, data exfiltration or destructive commands. An AI review of the same text found nothing harmful. It is MIT-licensed and scores 98/100 on trust signals. Skills are instructions an agent will follow, so read the file before installing it and do not approve commands you do not understand.
- Is product-showcase still maintained?
- The repository was last updated about 3 months ago. That is recent enough to be usable, but agent tooling moves fast, so check the instructions against your agent's current version.
Skill content
View source on GitHubname: product-showcase description: "Generate a comprehensive marketing website for a web app — multi-page with real screenshots, animated GIF walkthroughs, feature deep-dives, and workflow demonstrations. Browses the running app, captures screens and sequences, and produces a deployable site that actually teaches people what the product does. Especially useful for complex or agentic apps that are hard to explain. Triggers: 'showcase site', 'product page', 'show off the app', 'marketing site', 'demo site', 'product showcase', 'explain the app', 'how do I market this'." compatibility: claude-code-only
Product Showcase Generator
Generate a marketing website that actually teaches people what a web app does. Not just a hero and feature grid — a multi-page site with real screenshots, animated GIF walkthroughs of workflows, feature deep-dives, and progressive depth from "what is this" to "here's exactly how it works."
Especially valuable for complex apps, agentic AI tools, and anything where a static screenshot doesn't convey the value.
Depth Levels
| Depth | Output | Duration | |-------|--------|----------| | quick | Single page — hero, features, CTA. Same as before. | 15-20 min | | standard | Multi-page site — home, features page, how-it-works with screenshots. | 1-2 hours | | thorough | Comprehensive site — home, per-feature pages, animated GIF walkthroughs, use cases, comparison page, docs-style demo. | 3-6 hours |
Default: standard
Browser Tool Detection
Before starting, detect available browser tools:
- Chrome MCP (
mcp__claude-in-chrome__*) — preferred for authenticated apps - Playwright MCP (
mcp__plugin_playwright_playwright__*) — for public apps - playwright-cli — for scripted flows
Workflow
1. Gather Input
| Field | Required | Example |
|-------|----------|---------|
| App URL | Yes | https://app.example.com or http://localhost:5173 |
| App name | Yes | "Acme CRM" |
| Tagline | No | "The CRM that gets out of your way" |
| Target audience | No | "Small business owners" |
| Pricing info | No | Free tier, $29/mo pro |
| CTA text + URL | No | "Start Free Trial" → signup page |
| Testimonials | No | User provides or skip section |
2. Capture Screenshots
Use capture-screenshots (shipped in bin/) to capture the app. This is faster and more consistent than generating Playwright scripts each time.
Quick capture (all key pages at once):
capture-screenshots http://localhost:5173 \
--pages /,/dashboard,/contacts,/settings \
--output showcase/screenshots \
--prefix screen \
--mobile --dark
This produces desktop (1280x720), mobile (375px), and dark mode variants for each page in one run.
For authenticated apps:
capture-screenshots https://app.example.com \
--pages /,/dashboard,/settings \
--auth user:password \
--output showcase/screenshots \
--mobile --dark
What to capture:
a. First Impression — the main page/dashboard becomes the hero image. Note the immediate value proposition.
b. Features — each major section. Use --pages with all nav paths. Capture 6-10 key screens that tell the product story.
c. "How It Works" flow — the main workflow in sequence. Run capture-screenshots multiple times with --prefix workflow-step as you navigate through the flow steps.
d. Detail shots — zoom into specific UI elements. Use --full-page for scrollable content.
e. Both modes — --dark flag captures light and dark variants automatically. Use the best-looking mode for the hero.
Post-capture optimisation:
img-process batch showcase/screenshots --action optimise --max-width 1920 -o showcase/screenshots-opt
f. Extract the Value Propositions
Don't just list features. For each one, answer: why does the user care?
- BAD: "Contact management page"
- GOOD: "See every client, their history, and what needs attention — in one view"
- BAD: "Search functionality"
- GOOD: "Find anything in seconds — semantic search understands what you mean, not just what you type"
3. Generate the Site
Quick Mode: Single Page (same as before)
One HTML file: hero + feature grid + CTA. Use for MVPs and quick marketing.
Standard Mode: Multi-Page Site
showcase/
├── index.html # Home — hero, overview, feature highlights, CTA
├── features.html # All features with screenshots and descriptions
├── how-it-works.html # Step-by-step workflow walkthrough with screenshots
├── screenshots/ # All captured images
│ ├── hero.png
│ ├── feature-*.png
│ ├── workflow-step-*.png
│ └── *.gif # Animated walkthroughs
└── styles.css # Shared styles (or inline Tailwind CDN)
Home page: Hero with animated GIF or key screenshot, 3-4 feature highlights (not all features — just the best), "How It Works" summary (3 steps), CTA.
Features page: Every feature with a real screenshot and benefit-focused description. Group by category if there are 6+. Each feature gets enough space to actually explain what it does.
How It Works page: The primary workflow as a step-by-step visual guide. Each step has a screenshot (or animated GIF), a heading, and 2-3 sentences. This page answers "ok but what does using it actually look like?"
Thorough Mode: Comprehensive Site
showcase/
├── index.html # Home — hero, overview, value proposition
├── features/
│ ├── index.html # Feature overview grid
│ ├── [feature-1].html # Deep-dive: one page per major feature
│ ├── [feature-2].html # Each with screenshots, GIFs, use cases
│ └── [feature-n].html
├── how-it-works.html # Full workflow walkthrough
├── use-cases/
│ ├── [use-case-1].html # Scenario: "A day in the life of..."
│ └── [use-case-2].html # Scenario: "When a new client calls..."
├── compare.html # "Why [app] vs alternatives" (optional)
├── screenshots/
│ ├── hero.png
│ ├── feature-*/ # Per-feature screenshot sets
│ └── workflows/ # Animated GIFs
└── styles.css
Per-feature deep-dive pages: Each major feature gets its own page with:
- Hero screenshot of the feature in action
- "What it does" — 1-2 paragraphs explaining the value
- "How it works" — step-by-step with screenshots or GIF
- "Why it matters" — the problem this solves
- Edge cases or power-user tips
- Link to next feature (flow between pages)
Use case pages: Story-driven pages that show the app in a real scenario:
- "It's Monday morning. You open the dashboard and see..."
- Walk through a realistic workflow with screenshots at each step
- Show the outcome — what's different because the user used this app
- These are the most persuasive pages for apps that are hard to explain
Comparison page (optional): "Why [app] vs [alternatives]" — honest comparison, not marketing fluff. Feature table, key differentiators, who it's best for.
4. Animated GIF Walkthroughs
Static screenshots don't convey workflow. For key features, capture animated GIFs that show the actual interaction:
How to capture (using Playwright or Chrome MCP):
- Navigate to the starting state
- Start recording screenshots at ~2fps
- Perform the workflow (click, type, navigate)
- Stop recording
- Combine frames into a GIF
Generating the GIF — capture sequential screenshots then combine:
# Capture each step with a sequential prefix
capture-screenshots http://localhost:5173/clients \
--prefix workflow-01 --output .jez/screenshots
# ... navigate to next state ...
capture-screenshots http://localhost:5173/clients/new \
--prefix workflow-02 --output .jez/screenshots
# Combine frames into GIF (Python one-liner using Pillow)
python3 -c "
from PIL import Image; import glob
frames = [Image.open(f) for f in sorted(glob.glob('.jez/screenshots/workflow-*.png'))]
frames[0].save('showcase/screenshots/workflows/create-client.gif',
save_all=True, append_images=frames[1:], duration=500, loop=0)
"
What to animate:
- The primary "create something" flow (2-4 seconds)
- A search/filter interaction (show results appearing)
- A drag-and-drop or reorder operation
- Dark mode toggle (satisfying visual)
- Any "magic moment" where the app does something impressive (AI classification, instant search, real-time update)
GIF guidelines:
- Max 10 seconds / 20 frames — shorter is better
- Capture at 1280x720, display at 640x360 (half size for file size)
- Add a brief pause (3 frames) on the final state so viewers see the result
- Loop continuously — no "click to play"
- If the GIF would be >5MB, use fewer frames or crop to the relevant area
Display in HTML:
<div class="browser-frame">
<div class="browser-frame-bar">
<span class="browser-frame-dot"></span>
<span class="browser-frame-dot"></span>
<span class="browser-frame-dot"></span>
</div>
<img src="screenshots/workflows/create-client.gif"
alt="Creating a new client in 3 clicks"
loading="lazy" width="640" height="360">
</div>
5. Explaining Agentic / AI Apps
Agentic apps are especially hard to market because the value is invisible — the AI does work the user never sees. Standard screenshots show a chat interface. That's not compelling.
Patterns that work for agentic apps:
| Pattern | What it shows | Example | |---------|-------------|---------| | Before/after | What the user used to do manually vs what the agent does | "Used to: copy-paste from 3 systems. Now: agent does it in background." | | Timeline | What happens over time — show the agent working across hours/days | "8am: agent checks inbox. 9am: classifies 47 emails. 10am: flags 3 urgent." | | Result showcase | Skip the process, show the output | "Agent mined 1,200 emails → 89 clients, 340 contacts, 2,100 knowledge facts" | | Side-by-side | Show the agent's work next to what a human would have done | Split screen: left is the raw email, right is the extracted structured data | | Magic moment GIF | One animation of the most impressive thing | User asks a question → agent searches knowledge → returns answer with sources |
Copy tips for agentic apps:
- Lead with the outcome, not the technology ("Know every client's history" not "AI-powered CRM")
- Show volume ("Processed 1,200 emails" is more impressive than "Processes your emails")
- Use time comparisons ("What took 2 hours now takes 30 seconds")
- Avoid jargon ("Finds connections in your data" not "Semantic vector search with RAG")
5. Screenshot Presentation
Screenshots are shown in browser-frame mockups using CSS:
.browser-frame {
border-radius: 8px;
box-shadow: 0 25px 50px -12px rgba(0,0,0,0.25);
overflow: hidden;
border: 1px solid rgba(0,0,0,0.1);
}
.browser-frame-bar {
background: #f1f5f9;
padding: 8px 12px;
display: flex;
gap: 6px;
}
.browser-frame-dot {
width: 10px;
height: 10px;
border-radius: 50%;
background: #e2e8f0;
}
This gives screenshots a polished "app in a browser" look without needing to edit the images.
6. Site Navigation (multi-page)
Multi-page sites need consistent navigation:
<nav>
<a href="/">Home</a>
<a href="/features/">Features</a>
<a href="/how-it-works.html">How It Works</a>
<a href="/use-cases/">Use Cases</a> <!-- thorough only -->
<a href="#pricing">Pricing</a>
<button>Get Started</button>
</nav>
- Sticky nav on all pages
- Current page highlighted
- Mobile hamburger menu
- CTA button always visible in nav
7. Output
After generating, tell the user:
- Preview:
python3 -m http.server -d showcasethen openhttp://localhost:8000 - Deploy: drag the
showcase/folder to Cloudflare Pages, Netlify, or any static host - List any placeholder content that needs replacing
- Note which GIFs were generated and their file sizes
Design Patterns
Colour: If the app has a clear brand colour, extract it and use it as the primar
Truncated for display — read the full file on GitHub.
Related Skills
Agent-Reach
96.8kGive your AI agent eyes to see the entire internet. Read & search Twitter, Reddit, YouTube, GitHub, Bilibili, XiaoHongShu — one CLI, zero API fees.
Scrapling
87.1k🕷️ An adaptive Web Scraping framework that handles everything from a single request to a full-scale crawl! Don't be shy, join here: https://discord.gg/EMgGbDceNQ and follow here for daily tips and tricks: https://x.com/Scrapling_dev
ruflo
74.5k🌊 The original agent harness. Deploy intelligent multi-player swarms, coordinate autonomous workflows, and build conversational AI systems. Features adaptive memory, self-learning intelligence, federation, vector RAG integration, and native Claude Code / Codex / Hermes and many more Integrated
algorithmic-art
177.9kCreating algorithmic art using p5.js with seeded randomness and interactive parameter exploration. Use this when users request creating art using code, generative art, algorithmic art, flow fields, or particle systems.
Languages
Trust signals
From repository metadata: license, adoption, age and documentation. Not a code audit — see the Safety scan above for what the skill file itself contains.
