# Get an alert when a price drops, set up from ChatGPT or Claude

Source: https://clickruntime.com/use-cases/price-drop-alerts

> Tell your assistant which product and your target price. ClickRuntime checks it on a schedule, remembers the prices it has seen and emails you once when it drops, not every time it checks.

## Can ChatGPT or Claude track a price and tell me when it drops?

They can look a price up while you chat, but they can't keep checking afterwards or remember what the price was yesterday. With ClickRuntime connected, you ask once and it sets up an automation that reads the price on a schedule, stores each reading and emails you when it falls below your target.

### Why the assistant can't do this on its own

- A price drop is a change over time. Spotting one needs a check that keeps running and a record of earlier prices. A chat has neither.
- Built-in scheduled tasks re-run a prompt and reply in the chat. They don't keep a price history to compare against, so they either repeat the same answer or miss the drop.
- You want an alert once, not every two hours. That needs memory of whether you were already told, which lives in the automation's state on ClickRuntime.

## Set it up from your AI assistant

Prompt to copy:

> Create a ClickRuntime automation: every 2 hours, check the price on https://shop.example.com/products/monitor-27. Email me when it drops below $280. Tell me only once per drop, and include the lowest price seen so far.

1. **Connect ClickRuntime to your assistant.** Add ClickRuntime as a connector (MCP server) in ChatGPT, Claude, Cursor or another MCP client, using https://clickruntime.com/api/mcp. You sign in with Google or email; there is no API key to copy. This is done once per assistant.
2. **Paste the prompt.** Start a new chat and paste: “Create a ClickRuntime automation: every 2 hours, check the price on https://shop.example.com/products/monitor-27. Email me when it drops below $280. Tell me only once per drop, and include the lowest price seen so far.” Change the details (times, channels, addresses) to yours.
3. **Read the draft.** The assistant restates what will happen, asks for anything missing, such as your timezone or which channel to post to, and saves a draft automation. A draft never runs on its own, so nothing happens until you agree.
4. **Connect accounts and save keys.** A public product page needs no accounts. The email goes to the address you signed in with. If the shop has a price API that needs a key, the assistant sends you to the ClickRuntime dashboard to save it as a secret; it never asks for the key in chat.
5. **Test it.** The assistant runs it once and shows the price it read, so you can check it matches the page. It can also run it with a higher test target to show the email you'd get.
6. **Turn it on.** The assistant tells you what will run, when, and which services it will contact, then asks your permission to turn it on. From then on it runs on ClickRuntime's servers, and you can close the chat.
7. **Check it in the ClickRuntime dashboard.** Each run shows the price read and whether it was above or below target, for example “$319, above target”. State shows the lowest price seen and whether an alert was already sent for the current drop.

Connector setup for each assistant: https://clickruntime.com/assistants

## What the assistant writes

`price_drop_watch.py`

```python
import json
import re
from bs4 import BeautifulSoup
from clickruntime import on_trigger, state, notify, http, log

URL = "https://shop.example.com/products/monitor-27"
TARGET = 280.00

def read_price(html):
    soup = BeautifulSoup(html, "lxml")
    # Most shops publish the price as structured data for search engines.
    for tag in soup.select('script[type="application/ld+json"]'):
        try:
            data = json.loads(tag.string or "")
        except ValueError:
            continue
        offers = data.get("offers") if isinstance(data, dict) else None
        if isinstance(offers, list):
            offers = offers[0] if offers else None
        if isinstance(offers, dict) and offers.get("price"):
            return float(offers["price"])
    node = soup.select_one("[itemprop=price], .price")
    if node is None:
        raise RuntimeError("No price found on the page")
    return float(re.sub(r"[^0-9.]", "", node.get("content") or node.get_text()))

@on_trigger
def run(event):
    price = read_price(http.get(URL, headers={"User-Agent": "ClickRuntime price watch"}).text)
    lowest = state.get("lowest")
    if lowest is None or price < lowest:
        state.set("lowest", price)
        lowest = price

    alerted = state.get("alerted") or False
    if price < TARGET and not alerted:
        notify.email_me(
            f"Price drop: now {price:.2f}",
            f"{URL}\nNow {price:.2f}, your target {TARGET:.2f}. Lowest seen {lowest:.2f}.",
        )
        state.set("alerted", True)
        return {"summary": f"{price:.2f}, alert sent"}
    if price >= TARGET and alerted:
        state.set("alerted", False)   # back above target: alert again on the next drop
    log.info("Price checked", {"price": price})
    return {"summary": f"{price:.2f}, {'below' if price < TARGET else 'above'} target"}
```

- The price is read from the page's structured data first, which is steadier than the visible layout. A CSS fallback covers shops without it.
- Plain code compares the number with your target. No AI is needed for a threshold, so no AI tokens are used.
- The alerted flag in state makes sure you get one email per drop, and resets when the price goes back up.

## A schedule, not a trigger link

Use a schedule: shops don't send you a message when they change a price. Every few hours is enough for most products; flash sales may justify every 15 or 30 minutes. The shortest interval is every 5 minutes.

If you sell the product yourself and your store sends product-update webhooks, a trigger link is better than checking.

## Limits

- Every 5 minutes is the shortest interval.
- Pages are fetched without a browser. If the price only appears after JavaScript runs, ask the assistant to use the shop's product JSON or API.
- Some large shops block automated requests or require their own product API. Check the shop's terms.
- Prices can differ by country and currency. The automation sees the price the shop serves to a server in the cloud.

## Troubleshooting

### The run fails with “No price found”.

The shop changed its page or loads the price with JavaScript. Ask your assistant to “fix my price watch” and it reads the error and finds another source, often the product's JSON.

### The price it reads is in the wrong currency.

Ask the assistant to add the shop's country or currency parameter to the URL, or to convert with a currency API.

### I never got an alert even though the price dropped.

Check the runs in the ClickRuntime dashboard: the price read is in each result. If an alert was already sent for an earlier drop, it waits until the price goes back above target.

## Related

- [Shopify automations](https://clickruntime.com/apps/shopify)
- [Slack automations](https://clickruntime.com/apps/slack)
- [Google Sheets automations](https://clickruntime.com/apps/google-sheets)
- [watch-pricing-page](https://clickruntime.com/templates/watch-pricing-page)
- [Website change alerts](https://clickruntime.com/use-cases/website-change-alerts)
- [Monitoring and alerts](https://clickruntime.com/use-cases/monitoring-and-alerts)
- [Stateful automations](https://clickruntime.com/use-cases/stateful-automations)
- [Assistant scheduled tasks alternative](https://clickruntime.com/alternatives/chatgpt-claude-scheduled-tasks)
- [IFTTT alternative](https://clickruntime.com/alternatives/ifttt)
- [Docs: state](https://clickruntime.com/docs/state)

## Extend it: what you can add next

The first version is rarely the last. Because the automation is a Python script with memory, AI and any API available, you grow it by asking your assistant for the change. Each change is a new version, and a failed deploy never replaces the working one.

- **Track a list of products with a target price each** (Multi-step logic)
  Prompt: "Change my price watch so it checks these five product links, each with its own target price, and names the product in the alert."
- **Record every reading in a Google Sheet to see the price history** (Any app or API)
  Prompt: "Add to my price watch: add a row to my Price history sheet with the date, product and price on every check."
- **Alert on the lowest price in 90 days, not a fixed target** (Memory between runs)
  Prompt: "Change my price watch so it alerts me when the price is the lowest it has been in the last 90 days, instead of using a fixed target."
- **Send the alert to a Slack channel your household or team shares** (Any app or API)
  Prompt: "Change my price watch so it posts drops to #deals in Slack as well as emailing me."
- **Say what changed besides the price, such as stock or what's included** (Built-in AI)
  Prompt: "Add to my price watch: when the price drops, also tell me in one line if the stock status or what's included changed."
- **Check more often during sale weeks** (Schedules)
  Prompt: "Change my price watch to check every 15 minutes from 25 to 30 November, and every 2 hours the rest of the time."
- **Add products by sending a link to a trigger link** (Trigger links)
  Prompt: "Add a trigger link to my price watch so I can add a product by sending it a link and a target price, for example from a phone shortcut."

## Questions

### Can ChatGPT track prices for me?

Not by itself: it checks a price only while you chat and keeps no history. Connect ClickRuntime and ChatGPT sets up an automation that checks on a schedule, keeps the readings and emails you when the price drops below your target.

### Can Claude alert me when a price drops?

Yes, through ClickRuntime. Ask Claude with ClickRuntime connected; the automation runs on ClickRuntime's servers after the chat is closed.

### How often can it check a price?

Every 5 minutes at most. Every few hours suits most products.

### Will I get an email on every check?

No. The automation remembers whether it already told you about the current drop, so you get one email per drop.

### Can it watch Amazon?

Large marketplaces often block automated page requests and offer their own product APIs. The assistant can use such an API if you have access and save its key in the ClickRuntime dashboard.

### What can I add later?

More products, a price history sheet, Slack alerts or a lowest-in-90-days rule. The Extend it list on this page has prompts for each.
