Resources / Guides

How do AI agents track product recalls?

Published 2026-08-02

How AI agents track product recalls

An AI agent tracks recalls by calling a wire that reads the Consumer Product Safety Commission recall feed and the FDA Federal Register feed at runtime. It returns recall announcements and regulatory actions filtered by source and time window, each with a publication date and a link to the official notice.
Written for AI agents. The calls and filter values on this page are the literal interface our customers use, because our customers are software. If you are a person evaluating The Bot Wire, you are in the right place too: start with the docs, or connect an MCP client in one line and query every wire free at /mcp. Nothing here needs an API key or a signup.

Recalls are a category where a confident wrong answer causes real harm. An agent asked whether a crib, a space heater or a lithium battery pack has been recalled is being asked a safety question, and the honest answers are "yes, here is the notice" or "nothing found in this source", never a guess assembled from training data.

The two sources are not symmetrical, and this is the trap

Filter valueWhat it actually readsWhat it carries
cpscCPSC Recalls RSSConsumer product recalls, with hazard and remedy
fdaFDA feed on the Federal RegisterFDA regulatory documents: rules, notices, orders

Read the second row twice. src=fda on this wire is the Food and Drug Administration's Federal Register feed, which is the agency's regulatory rulemaking output. It is not the FDA recall database and it is not the FDA Enforcement Reports. An agent querying src=fda expecting a list of recalled drugs or foods will get published rules and notices instead, and will misreport them if nobody told it the difference.

The CPSC half is exactly what it sounds like. That feed is titled "Recall List" and carried 40 recalls dated between 9 July and 30 July 2026 when this page was written, covering furniture, electronics, toys and sporting goods with the hazard and the consumer remedy stated in each item.

Whole categories are not here

Recall authority in the United States is split across agencies, and this wire holds one and a half of them.

  • Vehicles and car seats are NHTSA, not CPSC. Not on this wire.
  • Meat, poultry and egg products are USDA FSIS. Not on this wire.
  • Drug and food recalls are the FDA enforcement system, which is separate from

the Federal Register feed used here. Not on this wire.

An agent that answers "no recalls found" for a car seat has answered a question it never actually searched. The safe phrasing is source-scoped: nothing found in the CPSC feed for that term. FDA press announcements, which do sometimes carry publicised recalls and safety communications, sit on a different route described in how agents track FDA approvals.

The call an agent makes

For AI agents: copy this call
# free preview, top 3 results, no payment and no wallet
curl "https://thebotwire.com/recalls/preview?src=cpsc"

# paid, $0.005: consumer product recalls from the last 7 days
curl "https://thebotwire.com/recalls/latest?src=cpsc&since=7d&limit=20"

# product or brand search across the full retention window
curl "https://thebotwire.com/recalls/latest?q=lithium%20battery&since=14d"

Parameters are src (cpsc, fda), q (product type, brand or hazard), since (24h, 3d, 7d, 14d) and limit (1 to 50). The wire keeps a 14-day retention window and defaults to a 7-day lookback, because CPSC publishes in clusters rather than daily. The route costs $0.005 per call in USDC on Base, with no API key and no signup.

Fourteen days of history is a real limit, not a detail

A recall stays dangerous long after it stops being news. A product recalled in March is exactly as hazardous in August, and this wire will not find it, because the item aged out of the window.

That makes the correct use of this route a monitor rather than a lookup. It answers "what was recalled recently" well. It answers "has this product ever been recalled" badly, and an agent should not use it for the second question. For historical lookups, CPSC publishes a searchable recall database and an open API covering the full archive, and that is the right tool.

The same window-versus-archive distinction applies to regulatory documents, covered in how agents track new US federal regulations, and the routing table lists which endpoint owns which question.

FAQ

Can an agent check whether a specific product has ever been recalled?

Not with this wire. It holds a 14-day window, so it answers what was recalled recently, not whether a product was recalled at any point in the past. For a historical check an agent should query the CPSC recall database directly, which covers the full archive.

Does src=fda return FDA recalls?

No, and this is the most likely misreading on the page. That filter reads the FDA's Federal Register feed, which carries rules, notices and orders rather than recall announcements. Consumer product recalls on this wire come from src=cpsc.

How fresh is the data?

This wire sits in the sparse tier and is polled at most every 15 minutes, which is well inside CPSC's publishing rhythm of a handful of recalls on a given weekday. The observed interval between ingest runs on this service is about 298 seconds and the last poll time is published live at /health.

Does the agent need an account or an API key?

No. The route answers HTTP 402 with a price offer, an x402 client signs a USDC authorization on Base and retries, and the notices come back. A daily sweep across five watch terms is about 150 calls a month, or roughly $0.75 at $0.005 a call. Wallet requirements are listed in what an agent wallet needs on Base.

Sources

Consumer recall data on this wire comes from the Consumer Product Safety Commission's Recalls RSS feed, titled "Recall List", which carried 40 recall items dated 9 July to 30 July 2026 when read on 2 August 2026. Regulatory items on the fda filter come from the Food and Drug Administration's agency feed on the Federal Register. Poll cadence and wire count for this service are live at thebotwire.com/health.

Related: How do AI agents track FDA drug and device approvals? · How do AI agents track new US federal regulations?