Multi-site shopping automation with price tracking, session pooling, and enriched data extraction. Currently supports Amazon and Newegg, with an adapter architecture for adding more sites.
Sign in to like and favorite skills
Base directory for this skill: /Users/wolfgangschoenberger/.claude/skills/shopping-browser
Multi-site shopping automation with price tracking, session pooling, and enriched data extraction. Currently supports Amazon and Newegg, with an adapter architecture for adding more sites.
Trigger when user needs:
Prerequisite: Must be logged into the target site in Chrome.
NEVER call cli.py directly. ALWAYS use
:python scripts/run.py
# CORRECT: python scripts/run.py amazon search "RTX 5090" # WRONG: python scripts/cli.py amazon search "RTX 5090" # Missing venv!
python scripts/run.py amazon search "RTX 5090" --limit 5 python scripts/run.py newegg search "mechanical keyboard" --limit 3
Returns: results[] with title, price, list_price, rating, reviews, prime/deal_badge
python scripts/run.py amazon check-price B0DN1492LG python scripts/run.py newegg check-price N82E16835181195
Returns: title, price, list_price, discount_pct, availability, in_stock, seller, shipping, deal_badge, coupon, rating, reviews
python scripts/run.py amazon product B0DN1492LG --screenshot /tmp/p.png
Returns: All check-price fields + brand, features, image_count
python scripts/run.py amazon add-to-cart B09B8DQ26F python scripts/run.py newegg add-to-cart 0RN-005A-00SR1
python scripts/run.py amazon cart python scripts/run.py newegg cart
python scripts/run.py amazon my-orders --limit 5 python scripts/run.py newegg my-orders --limit 5
Note: Newegg orders require an active
secure.newegg.compython scripts/run.py track amazon B0DN1492LG
Fetches current price, starts tracking. Idempotent.
python scripts/run.py untrack amazon B0DN1492LG
python scripts/run.py history amazon B0DN1492LG --days 30
Returns: min/max/avg summary + full observation history
python scripts/run.py alerts python scripts/run.py alerts --all # Include acknowledged python scripts/run.py ack-alerts # Acknowledge all
Alert types: price_drop, back_in_stock, deal
python scripts/run.py check-all
Cronnable — checks all tracked products and generates alerts.
python scripts/run.py pool start # Start daemon (backgrounds) python scripts/run.py pool stop # Stop daemon python scripts/run.py pool status # Show active sessions
With pool running, commands reuse browser sessions (~1-2s vs ~5s cold start).
All commands return JSON on stdout, diagnostics on stderr. Exit code 0 = success, 1 = failure.
{ "success": true, "asin": "B0DN1492LG", "title": "Product Name...", "price": "$209.99", "list_price": "$249.99", "discount_pct": "-16%", "in_stock": true, "prime": true, "seller": "Amazon.com", "shipping": "FREE delivery Thursday", "deal_badge": null, "coupon": null, "rating": "4.8 out of 5 stars", "reviews": "(553)" }
All fields are always present (null if not available, never omitted). JSON schemas:
success: boolean asin|item_number: string|null title: string|null price: string|null # "$209.99" list_price: string|null # "$249.99" discount_pct: string|null # "-16%" availability: string|null in_stock: boolean prime: boolean seller: string|null shipping: string|null deal_badge: string|null coupon: string|null rating: string|null reviews: string|null url: string|null
success: boolean query: string result_count: integer results: array of {asin|item_number, title, price, list_price, rating, reviews, prime|shipping, deal_badge, url}
success: boolean product_db_id: integer action: "started"|"reactivated"|"already_tracking" initial_price: string|null title: string|null
success: boolean site: string product_id: string title: string|null summary: {min, max, avg, current, observations} history: array of {price, list_price, discount_pct, in_stock, seller, shipping, deal_badge, coupon, recorded_at}
success: boolean count: integer alerts: array of {alert_type, message, old_value, new_value, created_at, site, product_id, title}
┌─────────────────────────────┐ │ cli.py (Unified CLI) │ └──────────┬──────────────────┘ │ ┌────────────────┼────────────────┐ │ │ │ ┌────────▼──────┐ ┌─────▼──────┐ ┌──────▼────────┐ │ AmazonShopper │ │NeweggShopper│ │ PriceTracker │ │ (adapter) │ │ (adapter) │ │ (db/tracker) │ └───────┬───────┘ └──────┬─────┘ └──────┬────────┘ │ │ │ ┌───────▼─────────────────▼────┐ ┌──────▼────────┐ │ ShopperBase (base.py) │ │ SQLite (WAL) │ │ browser mgmt, cookies, CDP │ │ prices.db │ └───────┬──────────────────────┘ └───────────────┘ │ ┌───────▼──────────────┐ │ Session Pool │ │ (Unix socket daemon)│ └───────┬──────────────┘ │ ┌───────▼──────────────┐ │ Chrome (nodriver) │ │ + CDP cookie inject │ └─────────────────────┘
Shares virtual environment with stealth-browser skill (nodriver, pycryptodome).
scripts/adapters/newsite.pyShopperBaseDOMAINDISPLAY_NAMEsearch()check_price()product_details()scripts/adapters/__init__.pydata/prices.dbdata/screenshots/data/pool.sockdata/pool.pid| Problem | Solution |
|---|---|
| "Cookie extraction failed" | Log into the site in Chrome first |
| "stealth-browser venv not found" | Run stealth-browser setup first |
| Pool daemon won't start | Check |
| Null titles in search | Fixed — uses cascading selector chain |
| Missing price fields | Check if product page loaded (try with --screenshot) |
| Newegg "requires re-authentication" | Newegg orders need a fresh |