Can I vibe code SearchTap?

Build a custom, high-performance ecommerce search and filter microservice using Node.js, Typesense, and React. 1. DATA MODEL & INDEXING: Define a product document schema containing id, title, description, handle, vendor, product_type, tags, price (float), compare_at_price, available (bool), inventory_quantity, created_at, variants (object array), and dynamic attributes (array of string key-values for facets). Write a background web-hook service that listens to product creation, update, and deletion events from Shopify/WooCommerce, normalizing catalog payloads and pushing them to Typesense within 500ms. 2. SEARCH & MERCHANDISING LOGIC: Implement a query parser endpoint exposed via a fast API edge worker. The endpoint must support term matching, configurable typo tolerance (1 typo for >4 chars, 2 typos for >8 chars), exact match attribute boosting (boost product title by 3x over description), and out-of-stock demotion (sort available=true higher). Add a simple JSON-based rule engine that checks for pinned products (e.g., query 'shoes' forces product_id '123' to position 1) and synonym replacements (e.g., 'tee' maps to 't-shirt'). 3. FRONTEND OVERLAY: Build a lightweight vanilla JS or React search overlay widget that injects into a store theme. Implement search-as-you-type debounce (150ms), recent search history cached in localStorage, instant auto-complete dropdown showing top 4 products with thumbnails + top 3 matching categories, and a full results drawer with multi-facet sidebar filtering (price range, vendor, tags). 4. FAILURE MODES: If the Typesense cluster drops, fall back gracefully to native platform REST API search with zero UI breakage. Log zero-result queries to a Postgres database for merchant review. Out of scope: vector search embeddings, complex multi-currency conversion, dynamic personalized reranking algorithms.

$ each button prefixes agent-specific run instructions · build your own product, never copy proprietary code, trademarks or designs