Skips product images, fonts and video while reading chart pages to cut proxy data (the main cost of a run). Also treats Amazon's "continue shopping" / robot-check pages (HTTP 200, no products) as a block and retries with a fresh session instead of reporting an empty chart. Memory now scales with the job: charts of about 80 or more products (or many URLs, or subcategory crawls) run on 4096 MB for speed; small jobs stay on 2048 MB. Results and prices are unchanged.
Adds the Movers & Shakers chart, rank change since your previous run (previousRank, rankChange, changeStatus, dropped-out list), 11 more marketplaces (21 total), friendly category names (Electronics on amazon.de now works), and a list of valid categories when a chart comes back empty. Prices and events are unchanged.
Bounds the automated quality-test prefill to one chart product and reconciles RUN_SUMMARY product charges from the event-specific charging manager instead of the cumulative ChargeResult count. Customer defaults, pricing, visibility, categories, and tasks are unchanged.
Initial release. Scrapes four Amazon ranking charts — Best Sellers, New Releases, Most Wished For, Gift Ideas (Most Gifted) — by category across 10 marketplaces (US, UK, DE, FR, IT, ES, CA, JP, IN, AU). Input via explicit chart URLs (chart type + marketplace auto-detected per URL) or via Chart type + Marketplace + optional category slug. PlaywrightCrawler scrolls each chart to load the full top 100 (Amazon lazy-loads ranks 31-100 on scroll), with ?pg pagination. Optional one-level subcategory crawl. Residential proxy default, geo-matched to marketplace. 18 flat fields per product optimized for MCP / AI-agent consumption. PPE: $0.005 per product + $0.00005 actor start. Movers & Shakers chart planned for a future version.
