How to get OpenFDA Recall and Enforcement Monitor in an AI tool

Query the public openFDA enforcement API for food, drug, or device recalls with Lucene search filters. A run saves one row per result.

Set up the Apify actor

Apify is where this run is collected. An actor is the program that collects the rows. A task is that program saved with your settings, so the next section can ask for the latest rows. You need a free Apify account.

01
Open OpenFDA Recall and Enforcement Monitor and click Try for free. Create an Apify account if you do not have one. The Free plan does not ask for a card.

You are on the input form. Keep the sample for this first run.

{
    "dataset": "food",
    "searchQuery": "status:Ongoing",
    "maxItems": 5
}

03
Leave the proxy setting as it is. Click Start and wait until the run says Succeeded.
04
Open Storage, then Dataset. You should see recallNumber, productDescription, status, classification.
05
When you want the real list, raise maxItems. The sample cap is only there so the first run stays small.
06
Go back to the input and choose Save as a new task. Name it after this list.
07
Open the task. Copy the task ID from the address bar. Excel, Sheets, Power BI, and the other guides ask for this ID.
08
On the task, open Schedules and add a weekly run if you want fresh rows. Each run pays the start charge and the charge for the rows. The amounts are in the price list on this page.
09
Open Settings, then API & Integrations, and copy an API token. Excel, Sheets, and the other guides put this token in a download link. Anyone with that link can download the rows, so treat it like a password.

How to set up in an AI tool

Use the task ID and the API token from Set up the Apify actor, above.

01
Open the API tab on OpenFDA Recall and Enforcement Monitor.
02
Copy the MCP config, or call the run endpoint with the API token from Set up the Apify actor, above.
03
Ask for recallNumber, productDescription, status, classification, reportDate. A schedule still lives on the task from Set up the Apify actor, above.

Schema

Examples come from the actor sample, not a live result.

Name	Description	Example
`recallNumber`	Recall number	`F-1234-2024`
`productDescription`	Product description	`Frozen vegetable blend 16 oz bag`
`status`	Status	`Ongoing`
`classification`	Classification	`Class I`
`reportDate`	Report date	`2024-03-15`
`country`	Country	`United States`
`sourceUrl`	Source URL	`https://api.fda.gov/food/enforcement.json?...`
`scrapedAt`	Scraped at	`2026-09-09T10:00:00+00:00`

recallNumberRecall numberF-1234-2024
productDescriptionProduct descriptionFrozen vegetable blend 16 oz bag
statusStatusOngoing
classificationClassificationClass I
reportDateReport date2024-03-15
countryCountryUnited States
sourceUrlSource URLhttps://api.fda.gov/food/enforcement.json?...
scrapedAtScraped at2026-09-09T10:00:00+00:00