| name | track-product |
| description | Track an Amazon or Walmart product URL and alert when price drops below a threshold |
| version | 1.0.0 |
| metadata | {"hermes":{"tags":["price","shopping","ecommerce","monitoring"],"category":"productivity","requires_toolsets":["terminal"]}} |
| required_environment_variables | [{"name":"PRICEPILOT_DIR","prompt":"Path to the pricepilot project directory","required_for":"running price tools"},{"name":"NIMBLE_API_KEY","prompt":"Your Nimble API key","required_for":"price scraping"},{"name":"CLICKHOUSE_HOST","prompt":"ClickHouse Cloud host","required_for":"price history storage"}] |
Track Product
When to Use
Invoke when a user shares an Amazon or Walmart product URL and wants price monitoring.
Trigger phrases: "track", "monitor", "watch", "alert me when under", "notify me when cheaper than"
Examples:
Procedure
-
Extract URL and threshold from the user's message.
- URL: any
https://amazon.com or https://walmart.com product link
- Threshold: dollar amount after "under", "below", "cheaper than" — if absent, default to 10% below current price
-
Check current price via terminal:
cd $PRICEPILOT_DIR && python tools/check_price.py "<url>"
Returns JSON: {"title": "...", "price": 99.99, "source": "amazon", "currency": "USD"}
-
Compute effective threshold: use user's value if given, else round(price * 0.9, 2)
-
Store price event in ClickHouse:
cd $PRICEPILOT_DIR && python tools/store_price.py "<user_id>" "<product_id>" "<title>" "<url>" "<source>" <price>
-
Register product for tracking:
cd $PRICEPILOT_DIR && python tools/add_tracked.py "<user_id>" "<product_id>" "<title>" "<url>" <threshold>
-
Reply to user:
"✅ Tracking [product name] — currently $[current_price]. I'll alert you when it drops below $[threshold]."
Pitfalls
- If the URL is a search results page (not a product), ask the user for the direct product URL
- If
check_price.py returns {"error": ...}, tell the user the URL couldn't be scraped and ask them to verify it
- Amazon ASIN is in the URL as
/dp/XXXXXXXXXX — use that as product_id
- Walmart item ID is the trailing number in the URL — use that as product_id
Verification
After tracking, confirm by running:
cd $PRICEPILOT_DIR && python tools/get_tracked.py "<user_id>"
Show the user their full tracking list.