Product Description Generator Agent Skill

Product Description Generator 📝

Generate platform-optimized product copy — titles, bullet points, descriptions, and backend keywords — for any major e-commerce platform. No API key required.

Installation

npx skills add nexscope-ai/eCommerce-Skills --skill product-description-generator -g

> For Amazon listings, use our dedicated skill with Cosmo algorithm optimization:

> `bash

> npx skills add nexscope-ai/Amazon-Skills --skill amazon-listing-optimization -g

> `

> See: amazon-listing-optimization

Two Modes

| Mode | When to Use | Input |

|------|-------------|-------|

| A — Create | Writing a new listing | Product specs + optional competitor URLs |

| B — Optimize | Improving existing listing | Current listing or URL + optional competitor URLs |

Both modes support competitor analysis — just include competitor URLs to enable it.

Supported Platforms

| Platform | Output Components |

|----------|-------------------|

| Amazon | Title (≤200) + 5 Bullets (≤500 each) + Description (≤2000) + Backend (≤250 bytes) |

| eBay | Title (≤80) + HTML Description |

| Walmart | Title (≤75) + Short Desc (≤150) + 10 Features + Long Desc |

| Shopify/DTC | SEO Title (≤60) + Meta Desc (≤160) + Product Description |

| Etsy | Title (≤140) + Description + 13 Tags (≤20 each) |

| TikTok Shop | Title (≤255) + Description (≤1000) |

| Lazada/Shopee | Title (≤120) + 5 Highlights + Description |

Usage Examples

Mode A — Create

Create a listing for my yoga mat on eBay UK.
Competitors: https://www.ebay.co.uk/itm/123456789, https://www.ebay.co.uk/itm/987654321
My product: 6mm TPE, non-slip, carrying strap included. Brand: ZenMat. Tone: Friendly.

Platform: Etsy. Product: hand-poured soy candle, lavender scent, 8oz glass jar, 40-hour burn time.
Target audience: gift buyers. Tone: Luxury.

Mode B — Optimize

Optimize this Shopify listing: https://mystore.com/products/portable-blender
Beat these competitors: https://amazon.com/dp/B09V3KXJPB, https://walmart.com/ip/123456

Find keyword gaps and rewrite this Etsy listing:
[paste current title, description, and tags]

Handling Incomplete Input

If user doesn't provide enough info, ask upfront:

To generate your listing, I need:

**Required:**
- Platform (eBay / Walmart / Shopify / Etsy / TikTok Shop / Lazada / Shopee)
- Product name and key features
- Brand name

**Recommended (better results):**
- 1-3 competitor URLs to analyze
- Target audience
- Tone preference (Professional / Friendly / Urgent / Luxury)

Which mode?
- **A — Create**: I'm writing a new listing from scratch
- **B — Optimize**: I have an existing listing to improve

💡 For Amazon listings, I recommend using [amazon-listing-optimization](https://github.com/nexscope-ai/Amazon-Skills/tree/main/amazon-listing-optimization) — it's optimized for Amazon's Cosmo algorithm.

Mode A Workflow — Create New Listing

Step 1: Collect Product Info

| Field | Required | Example |

|-------|----------|---------|

| product_name | ✅ | Portable blender |

| platform | ✅ | Etsy |

| brand | ✅ | BlendJet |

| key_features | ✅ | USB-C, 6 blades, BPA-free |

| specs | ✅ | 380ml, 175W motor |

| target_audience | 👍 | Gym-goers, travelers |

| use_cases | 👍 | Smoothies, protein shakes |

| competitor_urls | 👍 | 1-3 URLs to analyze |

| tone | Optional | Professional (default) / Friendly / Luxury / Urgent |

Step 2: Gather Keywords

If competitor URLs provided:

• Fetch each competitor page:

   Use web_fetch on each competitor URL.
   Extract: title, bullets/features, description, price, review count, brand name.

• If web_fetch fails or returns incomplete data:

   Fallback: web_search for "[product title from URL]" site:[platform].com
   Extract data from search snippets.

• Parse competitor content and extract keywords in these categories:
• Product-type terms: What it IS (yoga mat, exercise mat)
• Feature terms: What it DOES (non-slip, eco-friendly)
• Use-case terms: WHERE/WHEN used (home gym, yoga studio)
• Audience terms: WHO buys (beginners, athletes)
• Attribute terms: Specs (6mm, TPE material)

• Expand beyond competitors:

   web_search: "[product type]" best seller features what buyers want
   web_search: "[product type]" review complaints common issues
   web_search: site:[platform].com "[product type]"

If no competitor URLs provided:

• Discover keywords via web search:

   web_search: "[product name]" best seller [platform] features
   web_search: "[product name]" review what customers love
   web_search: "[product name]" vs alternatives comparison
   web_search: site:[platform].com "[product name]"

• Extract keywords from top 5 results following the same categories above.

⚠️ Critical: Remove all competitor brand names — never include them in output.

Step 3: Score and Prioritize Keywords

Score each keyword (1-9 points):

| Dimension | Scoring |

|-----------|---------|

| Frequency | In 3+ competitor titles = 3 pts / In 1-2 = 2 pts / Bullets only = 1 pt |

| Relevance | Core descriptor = 3 pts / Feature = 2 pts / Peripheral = 1 pt |

| Opportunity | Few competitors use = 3 pts / Most use = 2 pts / All use = 1 pt |

Assign to tiers:

🔴 Primary (7-9 pts)   → Title
🟡 Secondary (4-6 pts) → Bullets / Features
🟢 Tertiary (2-3 pts)  → Description
⚪ Backend (1 pt)       → Tags / Search Terms

Step 4: Generate Copy

Proceed to Generate Copy section.

Mode B Workflow — Optimize Existing Listing

Step 1: Analyze Current Listing

User may provide:

• Full listing copy (title, bullets, description pasted directly)
• Product URL (e.g., https://www.etsy.com/listing/123456)
• Platform + product identifier (e.g., "Etsy listing 123456")

If user provides URL or identifier only:

Use web_fetch on the provided URL.
Extract: current title, bullets/features, description, tags (if visible), price.

If web_fetch fails:

Fallback: web_search for the product title or identifier.
Ask user to paste the listing content manually if data is incomplete.

Once listing content is obtained, parse and extract:

• All keywords currently present
• Structure and format used
• Obvious gaps (missing features, weak benefits, no FABE structure)

Step 2: Gather Target Keywords

If competitor URLs provided:

Follow the same competitor analysis process as Mode A Step 2:

• web_fetch each competitor URL
• Extract their keywords
• Expand via web search

If no competitor URLs provided:

Discover ideal keywords for the product type:

web_search: "[product type]" top keywords [platform] 2024 2025
web_search: "[product type]" best seller features
web_search: site:[platform].com "[product type]" top listings

Step 3: Gap Analysis

Compare current keywords vs. target keywords:

## Keyword Gap Analysis

### ✅ Keywords You Already Have
| Keyword | Title | Bullets | Description |
|---------|-------|---------|-------------|
| yoga mat | ✅ | ✅ | ✅ |
| exercise mat | ❌ | ✅ | ❌ |

### ❌ Keywords You're Missing
| Keyword | Priority | Recommendation |
|---------|----------|----------------|
| non-slip | 🔴 High | Add to title |
| eco-friendly | 🟡 Medium | Add to bullet 2 |
| extra thick | 🟡 Medium | Add to bullet 3 |

Current Coverage: 12/20 keywords (60%)
Target Coverage: 90%+

Step 4: Rewrite

Generate optimized copy incorporating missing keywords.

Show Before → After for each component.

Proceed to Generate Copy section.

Generate Copy

Final step for all modes after keyword priority table is built.

Writing Framework: FABE

Apply to every bullet:

F — Feature:   What the product HAS or DOES
A — Advantage: Why this is BETTER than alternatives
B — Benefit:   What this MEANS for the customer
E — Evidence:  Spec, number, or proof that backs the claim

Lead with the Benefit — customers buy outcomes, not features.

Example:

❌ "Made with BPA-free Tritan plastic"
✅ "SAFE FOR YOUR FAMILY — BPA-free Tritan plastic means no harmful chemicals leaching into your smoothies, even after 1000+ uses"

Platform-Specific Writing Rules

Amazon (Cosmo Algorithm)

• Title: Brand + Primary Keyword + Attribute + Attribute + Secondary Keyword, ≤200 chars
• Bullets: [CAPS HEADER] — Benefit-led sentence with 1-2 keywords embedded
• Description: Hook → Features → Use cases → What's in box → CTA
• Backend: Space-separated keywords, no duplicates, ≤250 bytes
• ⚠️ Cosmo tip: Use situational language (when/where/why), cover multiple use cases
• 💡 For advanced Amazon optimization, consider amazon-listing-optimization

eBay (Cassini Algorithm)

• Title: Front-load exact-match keywords, 80 chars max
• Description: Repeat top 3 keywords naturally throughout, include specs table in HTML

Walmart

• Title: Brand + product name + primary attribute, ≤75 chars
• Short Desc: One-sentence value prop with primary keyword
• Features: 10 attribute-focused bullets, one fact per bullet

Shopify/DTC (Google SEO)

• SEO Title: Primary keyword + brand, written for Google (not just product name)
• Meta Desc: Keyword + benefit + CTA, drives CTR from search results
• Description: Storytelling structure with <h2>, <ul>, <strong> for on-page SEO

Etsy (Tag Matching)

• Title: Long-tail keyword phrase first, then style/material/occasion
• Description: Conversational, first 160 chars = meta description
• Tags: 13 tags (≤20 chars each), match phrases used in title exactly

TikTok Shop (Social Commerce)

• Title: Lead with problem or desire ("Tired of X?") → product → top feature
• Description: Hook → pain point → solution → 3 bullets → CTA. Conversational, emoji-friendly.

Lazada/Shopee

• Title: Brand + product + model + material + key attribute (completeness over cleverness)
• Highlights: 5 short bullets, one feature per line, spec-focused
• Description: Feature table + expanded use cases

Tone Guide

| Tone | Style | Best For |

|------|-------|----------|

| Professional | Authoritative, spec-focused, trust-building | Electronics, tools, B2B |

| Friendly | Conversational, benefit-focused, relatable | Kitchen, lifestyle, gifts |

| Urgent | Scarcity-driven, action words, problem-solving | Health, safety, seasonal |

| Luxury | Premium, sensory language, exclusivity | Beauty, fashion, premium goods |

Default: Professional if not specified.

Output Format

# ✅ Your Listing — Ready to Copy

Platform: [Platform] | Marketplace: [XX] | Tone: [Tone]

## Title
[title — copy directly into platform]

## Bullets / Features
1. [CAPS HEADER] — [text]
2. [CAPS HEADER] — [text]
3. [CAPS HEADER] — [text]
4. [CAPS HEADER] — [text]
5. [CAPS HEADER] — [text]

## Description
[description — copy directly into platform]

## Tags / Keywords
[keywords formatted per platform rules]

---

# 📊 Diagnostic Report

**Mode:** [A/B] | **Competitors analyzed:** [N] | **Keywords scored:** [N]

## Keyword Priority Table
| # | Keyword | Score | Tier | Placed In |
|---|---------|-------|------|-----------|
| 1 | [keyword] | 8 | 🔴 | Title |
| 2 | [keyword] | 6 | 🟡 | Bullet 1 |

## Keyword Coverage Map
| Keyword | Title | Bullets | Desc | Tags | Status |
|---------|-------|---------|------|------|--------|
| [kw] | ✅ | ✅ | ✅ | — | 🟢 |
| [kw] | ✅ | ❌ | ✅ | ✅ | 🟡 |

Coverage: X/Y keywords (Z%)
🟢 90%+ Excellent · 🟡 70-89% Good · 🔴 <70% Needs work

## ⚠️ Excluded Competitor Brands
[brands found in competitor copy — excluded from all output]

Integration with Other Skills

Looking for more e-commerce tools? Check out our other skill collections:

• Amazon Skills — Specialized tools for Amazon sellers: keyword research, listing optimization, PPC campaigns, sales estimation
• eCommerce Skills — Cross-platform tools for all e-commerce businesses

Limitations

This skill uses publicly available data via web search and page fetching. For real-time market data, exact search volumes, and advanced analytics, check out Nexscope.

Part of Nexscope AI — AI tools for e-commerce sellers.