FDA Recall Monitor: Drug, Food & Device Recall Alerts

Watch FDA recalls for the firms, brands, products or ingredients you care about — and get alerted only when something NEW is recalled. Each run pulls recent recalls from the keyless openFDA enforcement feeds (drug, food, device), keeps the ones matching your watch terms, and — because it remembers what it saw last time — flags just the new ones. Schedule it and point webhookUrl at Slack/Discord/email (via Make or Zapier) to turn it into a standing recall alarm. Works as a one-shot lookup too.

📋 Factual public records. Every row is an official FDA enforcement/recall report, reported verbatim (firm, product, reason, classification, dates). No scoring, no editorializing.

⚡ What makes it a monitor, not just a lookup

Most recall actors dump the feed every time. This one diffs against the previous run: on the first run it quietly sets a baseline (so you aren't spammed with history), and on every run after it surfaces — and webhooks — only recalls that are new since last time. Set a schedule once; hear about it only when it matters.

🔔 How alerting works

• webhookUrl — when a run finds new recalls, it POSTs { "newRecalls": N, "recalls": [...] } to your URL. Wire that into Slack, Discord, email, SMS via Make/Zapier, or your own endpoint.
• Schedule — add an Apify Schedule (e.g. every hour / every morning) so it keeps watch unattended.
• Apify can also email you on run events — but the webhook carries the actual new-recall payload.

An Actor runs in Apify's cloud, not on your machine, so there's no desktop pop-up — the webhook + a schedule is the cloud-native equivalent, and it reaches Slack/phone/inbox wherever you are.

📥 Input

{
  "watchTerms": ["peanut", "Acme Foods", "valsartan"],
  "productTypes": ["drug", "food", "device"],
  "classifications": ["Class I", "Class II"],
  "onlyNew": true,
  "webhookUrl": "https://hook.make.com/..."
}

Leave watchTerms empty to track all recalls in the window. classifications filters by severity (Class I = most serious). sinceDays sets the look-back (default 90).

📤 Example output (real run, trimmed)

A newly-detected recall for a watch on ["insulin"]:

{
  "isNew": true,
  "productType": "drug",
  "recallNumber": "D-0806-2026",
  "classification": "Class II",
  "status": "Ongoing",
  "recallingFirm": "Baxter Healthcare Corporation",
  "reason": "CGMP Deviations",
  "reportDate": "2026-09-09",
  "matchedTerms": ["insulin"]
}

The first run instead writes a baseline marker ("baseline": true, how many recalls are now tracked), so the run is self-explaining and no false "new" alerts fire for pre-existing recalls.

💰 Pricing (pay-per-event)

• Watch run — $0.01 per run (the fetch + diff).
• New recall alert — $0.02 per newly-detected recall. Quiet runs cost only the base; you pay the premium exactly when it catches something.

💡 Use cases

• Food brands / importers / grocers & marketplaces: know within the hour if a product or supplier you carry is recalled.
• Pharmacies / distributors / hospitals: watch drug & device recalls for your formulary or vendors.
• QA / regulatory / compliance: a standing Class I/II alarm across your brand and ingredient list.
• Agents: callable as an MCP tool — give it watch terms, get back only what's new.

🔑 Keyless & agent-callable

No API keys, no accounts, no paid child Actors — just public FDA (openFDA) enforcement data.

🚀 Ready-made tasks (one click)

• Food Recall Alert Monitor (FDA)

🤖 For AI agents (MCP)

https://mcp.apify.com/?tools=actors,docs,centorprat/fda-recall-monitor

Loads as centorprat--fda-recall-monitor. Claude Code:

claude mcp add apify -- npx -y @apify/actors-mcp-server --actors centorprat/fda-recall-monitor