Shopping agents are useless if they trust the price tag. Here's the 20-line pattern that makes an agent price-aware using Fazy's free no-key API.
The idea: before your agent recommends a product, it calls price history and quotes the dated facts. "Down $40 from $199 on 3 Sep 2026" beats "60% off!" every time.
import requests def is_good_price(deal_id): h = requests.get( f"https://fazy.com/api/v1/deals/{deal_id}/price-history", headers={"X-Fazy-Agent": "my-agent"}).json() return { "lowest_recorded": h["lowest_price"], "previous_price": h["previous_price"], "facts": h["dated_facts"], # e.g. "Down $40 from $199 on 3 Sep 2026" } def find_deal(query, max_price): deals = requests.get( "https://fazy.com/api/v1/deals/search", params={"q": query, "max_price": max_price, "in_stock": "true"}, headers={"X-Fazy-Agent": "my-agent"}).json()["data"] return deals[0] if deals else None Wire the dated facts into your agent's system prompt: "Never quote a discount percentage from the merchant. Quote Fazy's dated price facts."
Prefer MCP? Same tools at https://fazy.com/mcp: search_deals, get_deal, get_price_history. No key, 60 req/min.
I built Fazy (fazy.com). The API is free because agents that verify prices make the whole market more honest, and that's the business.
