filter_categories

com.pangolinfo/amazon-mcp · Verify this server

[Amazon category commercial-metrics filter] Filter categories by dozens of metrics (sales, GMS, search volume, conversion, return rate, price tier, competitor density, …) — or use as a "category detail" endpoint by passing a single categoryId. Use when: user says "find categories worth entering" / "high-sales categories" / "low return-rate categories" / "high search-volume but low competition categories" / "show me all metrics for category X"; category-level blue-ocean hunt; getting the 30+ metric snapshot of one category. Don't use: for niche-level (use filter_niches — finer granularity); for actual products in a category (use list_category_products); for just the readable name (use get_category_paths). Returns: data.items.data[{ id, categoryId, marketplaceId, timeRange, sampleScope, snapshotDate, unitSoldSum, glanceViewsSum, searchVolumeSum, netShippedGmsSum, buyBoxPriceAvg, buyBoxPriceTier, searchToPurchaseRatio, returnRatio, asinCount, offersPerAsin, newAsinCount, newBrandCount, avgAdSpendPerClick, unitSoldTrendDirection, unitSoldChangeRateBucket, ... trend + quantile-bucket fields }] + data.items.pagination.{ total, page, size, hasNext }. **Pagination**: use the 'page' param (default 1, 1-based, size capped at 10); 'pagination.hasNext=true' means more pages exist, 'hasNext=false' means last page. Pair with: ↑ required timeRange ('l7d' common) + sampleScope ('all_asin') + marketplaceId (defaults US); categoryId from search_categories / get_category_children; ↓ feed high-potential categories into list_category_products / list_bestsellers for real listings. Cost: ~1 point/page, ~5s. Tips: size capped at 10 (backend hard limit); only paginate when the user explicitly asks for more candidate categories — single-detail or quick-filter calls are fine on page 1; long-tail filter fields (unitSoldTrendDirections / metricChangeRateBuckets / dozens more) pass through via extraFilters.