Capafy logo Capafy
시작하기
E-commerce Listing Writer — Local Kit (Download)

E-commerce Listing Writer — Local Kit (Download)

by: LazyAI
Download package that runs on your own computer inside your AI agent (e.g. Claude Code or Codex — its subscription/API costs are yours). Requires Python 3.9+ with openpyxl. No third-party services or extra fees; works offline. Writes listings for Amazon, Etsy, Shopify, Shopee and momo from your product facts, checks platform limits and risky claims, and flags made-up specs. Exports Excel, CSV and a preview. AI-written; not legal advice.
#상거래#글쓰기#마케팅
평점
더 많은 평가가 필요합니다
판매
0
사용 방법
다운로드

E-commerce Listing Writer

What you're buying: a download package (skill instructions + a Python script) that runs on your own computer inside your own AI agent — it is not a cloud service.
You need: an AI agent that supports skills (e.g. Claude Code or Codex) — its subscription or API costs are yours; Python 3.9+ with openpyxl.
Third-party services: none. The script works offline; your agent writes the copy from the product facts you give it.

Product facts in — compliant listings for 5 marketplaces out, with no made-up specs.

Platforms and fields

Platform What it writes
Amazon Title (≤200, ≤80 recommended), 5 bullet points (CAPS LEAD: format), description, backend search terms (≤249 bytes)
Etsy Title (≤140), 13 tags (≤20 chars each), description (first 160 chars work as the search snippet)
Shopify Title, HTML body, tags, SEO title (≤70), meta description
Shopee Title (≤120), sectioned description, specs and box contents, up to 18 hashtags
momo (Taiwan) Short title (≤50), 3 feature points (≤40 each), formal spec description

Languages: English and Traditional Chinese (Taiwan usage); other languages work too.

Example

3 products (a magnetic power bank, a handmade stoneware mug, a linen tote) → 12 listings for Shopee (Chinese + English), Amazon (English) and momo (Chinese).

  • 6 errors were planted in the first draft, and the validator caught every one: "!" in an Amazon title, only 4 Amazon bullets, "free shipping" in a Shopee title, only 2 momo features, "Best" and "antibacterial" claims, and a made-up "20 hours" battery life (the product facts only said 20W).
  • After fixes: 12 listings, 0 errors, 0 warnings.
  • Output: listings.xlsx (one sheet per platform), one CSV per platform and language, and a card preview page.

The preview page — one product, three marketplaces, every card validated:

demo_preview_top.webp

Same facts, each platform's own format and length limits:

demo_preview_mug.webp

Try it — copy a prompt

  • "Here's my products.csv — write Amazon and Etsy listings in English, and flag anything that needs proof."
  • "Write a Shopify product page and SEO meta for this candle: 220 g soy wax, 45-hour burn, cedar & fig scent."

Why buy it

One payment, listings for your whole catalogue — no daily or monthly fees for a listing tool.

Who it's for

Marketplace sellers, Etsy makers, Shopify store owners, cross-border sellers and e-commerce assistants managing many SKUs.

Three automatic checks

  1. Platform rules: length limits, number of bullets and tags, banned title words ("free shipping", "sale", "Best Seller"…), Amazon's banned symbols, any word repeated more than twice.
  2. Risky or prohibited claims: medical claims (cure, weight loss, detox…), absolute guarantees (zero risk, lowest price…) and claims that need proof (antibacterial, organic, patented, eco-friendly…). English and Chinese lists, based on Amazon's restricted-claims policy, the FTC Green Guides and Taiwan's food, cosmetics and fair-trade rules.
  3. Made-up spec detection: every "number + unit" in the copy is checked against your product facts. If the facts say 20W but the copy says "20 hours", it gets flagged.

How it works

  1. Give the product facts (CSV, JSON or just tell your agent; the skill can create a blank template).
  2. Choose platforms and languages; the agent first builds keyword groups (main keyword + attributes, use cases, synonyms).
  3. The agent writes each listing, placing keywords where each platform weighs them.
  4. Validator → agent fixes → re-run, until 0 errors.
  5. Export to Excel, CSV and a preview page.

Output files

  • listings.xlsx — one sheet per platform, with a check-result column
  • <platform>_<language>.csv — UTF-8 with BOM, opens cleanly in Excel
  • preview.html — card preview marked PASS / CHECK / FIX, filterable by platform

Requirements

  • Python 3.9+ and pip install openpyxl
  • Any AI agent client (Claude, Codex, etc.)

Limitations (please read)

  • Copy is written by AI — check it before publishing.
  • It only uses facts you provide: if information is missing, the agent asks you instead of filling gaps.
  • The claim lists cover common risks but are not legal advice. Compliance with advertising law and platform policy is the seller's responsibility.
  • Platform limits change and can vary by category or marketplace — check your seller centre (rules live in an editable platform_rules.json).
  • Export columns follow common bulk-upload templates, but your account's official template may differ; paste the fields into the platform's own template.
  • It does not publish listings, scrape competitor prices or provide live keyword search volume.

Files

  • SKILL.md — agent workflow
  • scripts/listing_tool.py — template, validation and export
  • references/platform_rules.json — platform limits (editable)
  • references/banned_claims.json — risky and prohibited claims (English and Chinese)
  • references/writing-guide.md — keyword grouping and per-platform writing guide

View more