Daily buy-watch for your Airtable procurement wishlist

Build me an agent workflow that runs every weekday at 7am ET and does a buy-watch over my team's procurement wishlist in Airtable.

Trigger: cron, weekdays at 7:00 America/New_York.

Inputs I want to configure when I set this up: the Airtable base ID, the wishlist table name (default "Wishlist"), and the Slack channel to post buy-now alerts to. Assume the Wishlist table has these columns: Query (single-line text, the product search string), ProductId (single-line text, optional, the Google Shopping product_id), TargetPrice (number, USD), Requester (single-line text, a Slack handle like @jane), CurrentBestPrice (number, agent writes), BestStore (single-line text, agent writes), HistoricalLow (number, agent writes), Label (single-select with options: buy now, watching, skip; agent writes), Rationale (long text, agent writes, one line), LastChecked (date, agent writes).

Steps the agent should take on each run:

1. Call Airtable List Records on the Wishlist table to fetch every row. Treat any row missing Query as a skip.

2. For each row, get today's lowest current offer. If ProductId is set, call Real-Time Product Search Get Product Offers using that product_id and read the cheapest active seller. If ProductId is not set, call Real-Time Product Search Search Products with the Query, pick the best matching product from the top results, and use its top_offer (or call Get Product Offers on its product_id if you need the full seller breakdown). Capture the lowest price, the seller/store name, and the product URL.

3. For each row, call Real-Time Product Search Get Product Price History on the resolved product_id to get the historical low and median price across stores.

4. Use judgement to label each row:

- buy now: today's lowest price is at or below TargetPrice AND it is at or near the historical low (within a few percent).

- watching: trending downward (today is below the recent median) but still above TargetPrice, OR at target but well above the historical low.

- skip: above TargetPrice and either flat or trending up.

Write a one-line Rationale explaining the call in plain English, e.g. "$189 at Best Buy, 4% above all-time low of $182, target was $200".

5. For each row, call Airtable Update Record to write back CurrentBestPrice, BestStore, HistoricalLow, Label, Rationale, and LastChecked (today's date).

6. Collect every row whose new Label is "buy now" and post ONE consolidated Slack message via Send a Message to the configured channel. Format it as a short header ("Buy-now picks for {today}") followed by one bullet per item: the requester's Slack handle, the product name, the new price, the store, the deal link, and the one-line rationale. If there are zero buy-now items, post a short "No buys today, {N} items still watching" line instead of skipping silently, so the team knows the workflow ran.

Keep the Slack message consolidated, never one message per item. Make the run idempotent: if the same item is already labeled buy now from yesterday and the price has not changed, still include it in today's message (the team needs the daily nudge) but flag it as "still on sale".