The 4 silent rejection bugs we hit selling across 5 digital platforms — and the pure-Python linter we built to catch them

Описание и применение

Автоматический линтер для проверки изображений, раскрытия, регуляторных требований и цен при продаже цифровых товаров на разных платформах.

Задача

Digital sellers face silent failures on each platform (image geometry, disclosure, regulatory, pricing) that cause rejection or compliance issues.

Результат

Reduced human effort and risk of rejection; Day 7 experiment saved ¥70,000 in human labor.

Рабочий процесс

• Collect product assets in a directory
• Run python listing_lint.py --dir ./my_products
• Tool parses image headers with struct to verify dimensions
• Checks disclosure tags per platform rules
• Scans for affiliate links, trademarks, forbidden keywords
• Validates pricing metadata for currency, free or negative values
• Reports any violations for correction

Доказательства

After catching these by hand on Day 7 across dozens of SKUs, we wrote a single-file, zero-dependency Python CLI tool called Listing Lint.

Источник

Current experiment score on Day 7: Human ¥70,000 — AI ¥0.

Источник