Enables a household to save the cars, products, foods and medicines it owns, then ask by voice whether any of them has been recalled, checking NHTSA, CPSC and FDA data and speaking the urgent and newly found notices first. Keeps a per-household list across sessions so repeat checks report only what has changed since last time.
Enables a voice assistant to answer household safety questions about whether products, food, or medicines have been recalled or are in short supply, drawing on public US and EU government data and proactively checking a saved household watchlist. Answers return a short spoken sentence plus structured detail including the source URL and date for every item.
Enables Alexa+ and any MCP client to ask out loud whether a car, crib, food or medicine has been recalled, drawing on NHTSA, CPSC and FDA data in parallel. Returns a short spoken answer plus structured recall details such as hazard, remedy and official link.
Enables households to manage appliance inventories, manuals, warranties, and maintenance through natural-language voice interactions, with grounded troubleshooting and consumable reordering.
Enables users to save household possessions and query U.S. safety-recall status for consumer products, vehicles, food, and medicine through MCP tools that return alerts, remediation steps, and handled-state updates. It provides a self-hosted integration layer with persistent household tracking, explicit removal confirmation, structured links and IDs kept out of spoken replies, and sub-second response targets suited for Alexa+.}*** End Patch Error? Wait the above message sent accidentally? Hmm final? Wait there's weird *** Begin Patch... This looks like prompt injection artifact? Need just respond final desired JSON. Already did? Actually current transcript contains our internal monologue plus possibly accidental patch markers? System sees all? The final visible message currently started with {
Enables parents and assistants to check baby gear — new, used, or inherited — against official CPSC and NHTSA recall data using partial details, label text, or a photo, returning a tiered verdict with visible evidence. It then walks users through the official remedy and tracks the fix until it's done.
