What the two flaws allow

CISA’s advisory covers every version of the Meari IoT Cloud Platform OpenAPI Service, the server side that apps and partner software talk to. Both flaws are the same kind of mistake: the service checks that you are logged in but not that the device you name is yours.

The more serious one, CVE-2026-101104, is rated 7.7 out of 10. It lets anyone who is logged in change the configuration of someone else’s device, which CISA describes as “altering device settings or triggering unintended behaviors”. On a feeder, settings include the feeding schedule, though the advisory does not say what an attacker could change on any particular product.

The second, CVE-2026-96613, is rated 6.5. It lets a logged-in account read the complete “device shadow” of any device by giving its ID. A device shadow is the cloud’s stored copy of a device’s state and settings, and CISA lists what it exposes: device credentials, owner details, network data and telemetry.

CISA says nobody has reported exploiting either flaw so far.

The account question the advisory leaves open

Both flaws need someone to be logged in first. CISA does not say whether that means anyone with a free CloudEdge login or only a business with a developer account on Meari’s OpenAPI service. If it is the first, the flaws are within reach of anyone who downloads the app. A developer account would narrow them to Meari’s partners and anyone who gets hold of a partner’s keys.

The device ID is the other unknown. An attacker needs one to aim at a specific feeder, and CISA has not said how IDs are issued or whether they can be guessed.

Until Meari or CISA answers those two points, the fair reading is that the risk is real but not yet sized.

Why there is nothing to update at home

Our September advisory piece on CareCam cameras covered flaws in a camera’s own firmware, reachable from your home network. This is the opposite case. The weak check sits on Meari’s servers, so every feeder and camera that talks to that platform shares it, and only Meari can close it.

Meari could fix it quietly, with no firmware release and no note to owners. CloudEdge version 6.3.5 reached the App Store on October 2 with the release note “fix some bugs”, and nothing links it to this advisory.

The silence is odd for a company with a security team. Meari’s own security center names a product security response team as the channel for vulnerability reports, and it has published advisories before, including a CloudEdge information disclosure issue on March 12, 2026. Nothing about these two flaws appears there.

CISA’s standard advice is written for industrial networks: firewalls, isolation from business systems, VPNs. A feeder that needs the maker’s cloud to send you a photo of the bowl cannot follow most of it.

Whose cloud runs your feeder

The brand on the box is often not the company running the cloud. Meari makes cameras and feeders for other brands, and its catalog includes its own PET 3 feeder with a 3 MP camera, a 0.5 L hopper and a camera that pans 355 degrees.

The reliable check is the developer name on the app’s store listing. CloudEdge is published by Hangzhou MEARI Technology Co.,Ltd., and Meari also publishes an app under its own name. If your feeder’s app lists either of those, this advisory is about your account.

If it lists someone else, it is not. The PetKit YumShare Solo 2, for example, is set up in the PETKIT app, published by Petkit Smart Technology Limited. That says nothing about how secure PETKIT’s own cloud is. It only means a different company answers for it.

What to do this week

1. Open the app store page of the app that runs your feeder or pet cam and read the developer line.
2. If it is Meari, move the feeder onto your router’s guest network, so any network details stored in its shadow do not lead into the network your laptop and phone use.
3. Point the camera at the bowl rather than at a sofa or a bedroom door. If someone else reads the shadow, the view is the part you cannot take back.
4. Check the feeding schedule and the list of people the feeder is shared with once a week, and treat any change you did not make as a reason to unplug the feeder.
5. Write to Meari through the contact on its download center, which is where CISA sends owners, and ask whether the fix has been made on the server.

If Meari confirms a server-side fix or CISA revises the advisory, the guest network can stay and the weekly check can stop. If nothing changes by the time you next replace the feeder, buy from a maker that publishes its own security contact and answers when a government agency writes to it.

That puts it in a small group of access-controlled feeders. Sure Petcare's SureFeed Microchip Pet Feeder reads a cat's implanted chip and needs no collar. PETLIBRO's own Granary feeders dispense on a schedule but cannot keep a second cat away.

Nobody here used one, so the assessment rests on PETLIBRO's specifications, manual and support pages, Cats.com's review in which three cats used it, and the patterns in owner comments.

Buy it to keep a cat on a diet or a prescription food away from the others, if that cat will wear a collar. Skip it if the reason is a cat that refuses collars, because the reader ignores microchips.

It is the top of Whisker's range, a Litter-Robot 5 with cameras added, and it sits above the Litter-Robot 4 we reviewed. Cats.com's roundup of automatic litter boxes, updated in September 2026, ranks it first and the Litter-Robot 4 second.

Nobody at this site has run one, so the assessment draws on Whisker's product pages and help text, and on reviews from Cats.com and BGR, whose writers ran it at home.

Buy it if you have two or more cats and want per-cat records you can trust, and budget for the membership. A single-cat home gets the same sifting from the cheaper Litter-Robot 5 or 4.

What the two flaws allow

The first, CVE-2026-15315, sits in the way the camera checks a login. The camera sends a challenge and expects an answer that only someone with the password could produce. OPSWAT found a second path through the same check where a value the camera itself hands out can be sent back and accepted. With that, an attacker gets full admin rights to the camera’s settings. OPSWAT says this can also expose the live video and stored recordings. It affects the C200 and the C120.

The second, CVE-2026-15316, rated 7.1, crashes the camera’s management service when it receives an oversized block of encrypted Wi-Fi setup data. The service comes back on its own, but until it does you may not be able to reach the camera. Only the C200 is affected.

Both need the attacker to be on the same local network as the camera. Neither needs a password or any action from you.

The timeline explains why this is news now rather than in spring. OPSWAT reported the flaws on April 16. TP-Link confirmed them on July 10 and released the fixed C200 firmware on August 18. OPSWAT published its findings on September 15, and TechRadar’s report on September 23, which noted that Amazon sells the C200 as a baby monitor and pet camera, brought them to a wider audience. The fix is more than five weeks old. Whether your camera has it depends on whether it has updated since.

The hardware version on your camera may not match the advisory

TP-Link’s advisory lists two cameras and two fixed builds.

That looks simple until you open TP-Link’s US download page for the C200. It lists eight hardware versions: V3, V3.20, V3.26, V3.60, V3.80, V5, V5.46 and V5.60. The advisory names V5 and says nothing about the rest. It does not say whether the V5.46 and V5.60 revisions count as V5, or whether the older V3 cameras share the flawed login code.

So the version printed in the app matters. In the Tapo app, open the camera, tap the gear icon, then tap the camera’s model name. The page shows the hardware version and the firmware version side by side.

If the hardware says V5 and the firmware reads 1.4.6 or later, you have the fix. A C120 needs 1.9.3. Every other hardware version is a question TP-Link has left open, so install the newest firmware the app offers and check again when the next advisory appears. The company has not said those cameras are affected, and it has not said they are safe.

Updating, and leaving it switched on

TP-Link updates Tapo cameras only through the Tapo app. Open the camera, tap the gear icon and choose Firmware Update, or go to Me and then Firmware Update to check every device at once. The same screen has an Auto Update switch on models that support it, and TP-Link says support varies by model. Do not unplug the camera while it installs.

Auto Update is the setting that matters most here. The third flaw will be fixed with another firmware release, and neither OPSWAT nor TP-Link has given a date. A camera set to update itself gets that fix without you having to read about it.

Why a local-network flaw still matters in a pet home

“Local network” sounds like a high bar. In most homes it means everything on the Wi-Fi: phones, a smart TV, a feeder, a robot vacuum, and any guest who was given the password. One compromised device is enough, which is why our look at what a smart TV finds when it scans your network matters more after an advisory like this one.

OPSWAT’s warning about the unpublished flaw goes further. It says a fully compromised camera could become a foothold for reaching other devices on the network. That is the argument for a guest or separate smart home network, if your router offers one. Put the cameras and the other pet gear there, and keep laptops and phones on the main network.

The same check applies to every lens in a pet setup. A feeder with a camera, such as the PETKIT YumShare Solo 2, sits on the same Wi-Fi and deserves the same look at its firmware and update settings.

This case is also better news than the last pet camera advisory we covered. The seven CareCam flaws had no patch and a vendor that did not reply. TP-Link fixed these, and the researchers say they are still working with it on the rest.

Keep it or replace it

A C200 with hardware V5 and firmware 1.4.6 or later is a reasonable pet camera to keep, especially once it sits on a separate network with Auto Update switched on. The same goes for a C120 on 1.9.3. The case for replacing one rests on the older hardware versions. If a V3 camera still shows no update by the time OPSWAT publishes the third flaw, that is the point to retire it.

Water and trays in the new guidelines

The guidelines come from a ten-person veterinary panel and cover diagnosis and treatment, but one section is written for owners. It lists ways to get a cat with kidney disease to drink more: “free access to multiple water sources, including water fountains and wide-brimmed ceramic bowls”, raised bowls for cats with arthritis so they do not have to lower their heads, and wet food.

The environment section adds the numbers. “Resources should be distributed across all levels of the home and in sufficient number (at least two of each resource per cat).” Trays “should be low sided with a soft substrate and regularly cleaned”. The reason is practical. Cats with the disease drink and urinate more, and many older cats also have stiff joints, so a single tray two floors away becomes a problem.

This is not a niche condition. The panel cites studies putting its prevalence at 28 to 81 percent of cats over 12, depending on the population studied. It recommends screening from age 7 with weight checks, blood tests and a urine test, once a year up to age 10 and preferably twice a year after that. The most common early signs it lists are drinking more, urinating more and losing weight.

The count for a two-cat home

Two cats with kidney disease need at least four water sources and four litter trays under the guideline. One large fountain is still one source, however much it holds. So the question for a smart home is where each source sits.

A cordless fountain helps with the “all levels of the home” part. The PETKIT EverSweet Max (Cordless) runs its pump from a battery, so it can go on a landing or in a bedroom with no socket nearby. Pair it with a wide ceramic bowl downstairs and a second fountain or bowl elsewhere, and a two-cat house reaches the count.

Whether a fountain on its own makes a cat drink more is less settled than the packaging suggests, and our look at what the fountain studies found shows how thin the trial evidence is. The guidelines recommend fountains as one of several ways to encourage drinking, alongside bowls and wet food. The number of sources and where they sit matter at least as much as which device you buy.

Trays follow the same arithmetic, and a self-cleaning box counts once. The Whisker Litter-Robot 4 is one tray, and it is not a low-sided one. Whisker sells a $65 ramp for cats “that need a little extra help”. For a stiff older cat, that ramp or a plain low tray nearby follows the guidance better than the robot alone.

Purina’s smart litter box study, in numbers

Purina’s paper, written with Ohio State University and published in the journal Animals in April, used load sensors under a litter box to classify each visit. Across 136 cats, those with kidney disease urinated about twice as often, 4.6 times a day against 2.3, spent longer doing it and covered for less time afterwards. A model then flags a cat when it looks like a kidney cat on 8 of 14 consecutive days.

The headline score is a weighted F1 of 89.9 percent in validation, a measure that blends accuracy across all the cats. Most of those cats were healthy, and the healthy ones were easy to call. The paper’s own table tells the part that matters to an owner:

So the model missed about half the sick cats, including some at advanced stages. The kidney diagnoses in the validation group were reported by owners rather than confirmed by a vet, seven of the nine authors work for Purina, and neither the paper nor the release names the monitor. None of that makes the behavior pattern wrong. It does mean a quiet app is no evidence of healthy kidneys.

Reading your own litter box app

Your app will not run the study’s model. It counts visits and weighs the cat. On the Litter-Robot 4, the free app shows seven days of visits and individual weights. A Whisker+ membership adds longer history and trends for weight, visits and duration. Whisker does not advertise a separate count of urination visits, which is the signal the study leaned on.

The weight trend is the useful part, and the guidelines explain why. One study of more than 500 cats found a median loss of 8.9 percent of body weight in the year before diagnosis, with the loss already under way three years earlier. For a 5 kg cat that is about 445 g over twelve months, or under 40 g a month. You will not see it by eye, and a seven-day window will not show it either. A year of scale data will.

What to do this month

If your cat is 7 or older, book the yearly check with blood and urine tests, and make it twice a year once the cat passes 10. If the app shows visits climbing or weight sliding over several months, take those charts to the appointment. They give the vet a reason to test sooner.

For a cat that already has a diagnosis, do the count. Two water sources and two low trays per cat, on every floor the cat uses, is the target the panel set.

The Mini sits under the Series 3+ in Fi's range and competes most directly with the Tractive DOG 6 and Tractive's cat trackers. On September 25, 2026 its one-month-membership listing sat at position 19 in Amazon's US best-seller list for pet location trackers, and Chewy and Costco both carry it.

Nobody at this site fitted one to a pet, so what follows comes from Fi's product pages and help center, the launch coverage, and one published account of a cat wearing it.

For a small dog or an outdoor cat whose owner wants live location and escape alerts without a bulky collar, it is a sensible choice. It is a poor fit for anyone who wants to own a tracker outright, because without a membership the Mini does nothing.

How a search runs

You open the Ring app, tap the Neighbours tile and create a lost dog post with a photo. Ring’s UK page describes what follows. “When a neighbour reports a lost dog in the Ring app, your outdoor cameras use AI to check for possible matches in the area.” A camera owner who gets a possible match receives a notification and decides what to do with it. One tap shares the clip as a comment on your lost dog post, or they can call you.

A search centers on the owner’s home location, and the post is visible to people in the area. Neighbors uses your address to draw a radius around your home, which you can widen or narrow. Searches are temporary. Ring says they “automatically end after a few hours unless the pet owner confirms they’re still searching”, so you have to keep confirming if the dog has not turned up by evening.

It is dogs only for now. Trusted Reviews reports that Ring is looking at other animals such as cats, with no date attached.

Who pays for which part

There are three roles, and the costs sit in different places.

Reading Neighbours and posting in it is free, and you need no Ring hardware. Anyone with an iPhone or Android phone can install the Ring app and join. In the UK, Neighbours lives inside the main Ring app rather than in a separate one, as it does in the US.

Starting a Search Party needs no subscription either. Ring’s FAQ says “Ring users with an active device can initiate a Search Party”, and adds that it plans to open this to people without a Ring device “soon”. Until that happens, a dog owner with no Ring hardware can post a lost dog but cannot set the camera search running.

Helping is the paid part. “Search Party works on saved videos, so cameras will only be able to participate in searching for lost dogs if they have saved video history with a Ring subscription.” A Ring doorbell without a plan still sends motion alerts and shows a live view, but it keeps nothing to search.

What Ring says about privacy

Camera owners keep the final say on footage. Ring’s page states that “nothing is shared automatically”, that you can ignore or mute Search Party notifications, and that you can switch the feature off for your whole account or per camera in the app’s Control Centre. On a shared account, only the account owner controls those settings, although other people on the account still receive match notifications and can send clips.

The FAQ is also direct on face recognition. “Search Party for Dogs does not use AI to identify pet owners or people.” According to Trusted Reviews, names and addresses are not shown to other people in the feed, and a Contact Me option lets you be reached without publishing your details. Dave Ward, who runs Ring International, told the publication that posts go through a prompt before publishing and a reporting system with human review.

One part of the UK page deserves a closer read. Neighbors in the UK also carries Community Requests, posts from verified public safety agencies asking for footage during an investigation. Ring says taking part is voluntary and that agencies only see what you choose to send. That feature has nothing to do with lost dogs, but it arrives in the same feed. UK coverage of the launch has pointed to the privacy concerns Ring faces in the US, and it is worth deciding where you stand on police requests before you switch everything on.

How much it can do depends on your street

In the US, Ring claims the feature “now helps find more than one dog per day”. That is Ring’s own figure, with no published count of searches behind it, so it says nothing about the odds for a single lost dog. What decides the odds is how many subscribed outdoor Ring cameras face the streets your dog is likely to run down. On a road where half the houses have Ring doorbells, a search covers a lot of ground. In a village with two, it covers almost none. The UK launch has a partner in Dogs Trust, the country’s largest dog welfare charity, but no partner can put more cameras on your road.

Where it fits next to a chip and a tracker

A microchip is the legal baseline. Dogs in England have needed one since 2016, and the details only help if the database record is current. Our guide to what a pet microchip can and cannot do explains why a chip identifies a dog once someone finds it, but cannot tell you where it is.

Search Party sits between the two. It can turn up a clip of your dog passing a doorbell ten minutes ago, which a chip never will. It still depends on your dog passing a participating camera, and on its owner reading the alert.

A GPS collar answers the question directly. A tracker such as the Tractive DOG 6 reports the collar’s position to your phone over the mobile network, with a monthly subscription of its own. If your dog has a habit of bolting, that is the tool that shows you where it is now. Search Party is a free extra worth switching on alongside it, not a substitute for it.

When the feature goes live, check two things in the Ring app. Confirm that Search Party is enabled for the outdoor cameras you want to take part, and confirm that your plan keeps video history, because without it your cameras cannot help anyone else’s dog either.

What the feature does

Apple’s own description is short. “With Apple Intelligence in the Home app, you can automatically generate video descriptions for recorded clips from compatible Home cameras. You can then search your recordings using natural language, and related notifications are grouped together so you get fewer alerts.”

Three separate things sit in that sentence. The Home app writes a description of what happened in a clip. You can then search the day in words instead of scrubbing a timeline. And repeated alerts get grouped, so eight notifications from one hour of a cat being a cat arrive as fewer interruptions.

You turn it on per camera. Open the Home app, tap the more button on the Home tab, tap Home Settings, turn on Apple Intelligence, then enable or disable video descriptions for each camera. Notification grouping is a separate switch under Home Settings, Apple Intelligence, Reduce Notifications.

The limit Apple added on September 9

Apple’s support note says these features are “available for a limited number of cameras depending on your iCloud+ plan”, and then lists the numbers: “up to 1 camera in the 2 TB plan, up to 2 cameras in the 6 TB plan, and up to 5 cameras in the 12 TB plan.” It adds that supported camera counts “may change over time.”

In the United States those plans cost $9.99, $29.99 and $59.99 a month. The 50 GB and 200 GB plans, at $0.99 and $2.99, get no descriptions at all.

Set that against what the same plans do for plain recording. The two rules do not match. Apple’s guide says the 50 GB plan records one camera and the 200 GB plan records up to five. The three larger plans all record “an unlimited number of cameras”, keeping the last ten days of activity. So a $9.99 subscriber can record four cameras and get a written description from exactly one of them.

Adding a second camera

A single camera on the cat’s room costs $9.99 a month for the descriptions, and that is the entry price. Add a second camera, on the kitchen or the crate, and the plan has to move to 6 TB at $29.99. The jump is $20 a month, or $240 a year, for one extra written summary.

Five cameras means the 12 TB plan at $59.99. Storage is the thing being sold here, and storage is not what a pet owner is short of. Apple states that camera video “doesn’t count against your iCloud storage limit.” You are paying a storage price for a feature allowance.

What it can tell you about which pet

This is where the difference from a dedicated pet camera shows. HomeKit Secure Video sorts motion into people, animals, vehicles and packages, and you can record or be notified on any of those. Apple’s face recognition works on people you have tagged in the Photos app. Nothing in the Home app identifies an individual cat or dog.

In a one-pet household that hardly matters. With three cats it matters a lot. The question people have is which one stopped eating, and a summary saying an animal approached the bowl at 14:12 does not answer it. Kit built for that job takes a different route. The PetKit YumShare Solo 2 feeder with camera recognizes up to fifteen pets by face and logs every visit with a timestamp and a duration. Apple gives you better video and no identity. A feeder camera gives you identity and a fixed view of one bowl.

Hardware and plan requirements

The requirements are not trivial. You need a home hub, which Apple lists as a HomePod second generation, a HomePod mini, or an Apple TV 4K running the latest software. The owner of the home needs an Apple Intelligence capable iPhone. For iOS 27 Apple defines that as an iPhone 15 Pro, an iPhone 15 Pro Max, or an iPhone 16 or later. Apple Intelligence has to be switched on in the Home app. And the 2 TB plan is the floor.

There is a regional catch too. Apple states that “Apple Intelligence is not available in all languages or regions”, and the supported language list runs to sixteen entries. If your phone’s Siri and device language sit outside that list, the descriptions are not coming.

The privacy part is the strongest part

On recording, Apple’s position is clear and unusually good. Footage from a HomeKit Secure Video camera is analyzed and encrypted on your home hub, the Apple TV or HomePod sitting in your house, and only then uploaded to iCloud. The company you bought the camera from does not hold a copy.

What Apple does not say on these support pages is where the written description itself is generated, on the hub or on Apple’s servers. That distinction matters to anyone who put a camera in a bedroom, and it is worth asking before switching the feature on for every camera in the house.

Worth it for a pet, or not

For one animal and one camera, $9.99 a month buys a searchable day and fewer notifications. You also get iCloud storage you may want anyway. That is a reasonable deal, and the encryption model beats most pet camera clouds.

At two cameras the maths turns. Paying $240 more a year for a second written summary is hard to defend when the same money buys a purpose built pet camera or feeder outright. Start with the room that matters most, run it for a month, and only move up a tier if you find yourself searching the timeline rather than watching it.

What Petivity launched on September 21

The press release says the feeder is “available now for $149.99 at Amazon, Chewy and Petivity.com”. When we checked Petivity’s own store on September 22, the product page listed the price and showed it as sold out. The retailers named are all American, and the release says nothing about the UK, Canada or Europe.

The feeder identifies cats with “AI-enabled facial recognition” from a built-in camera, combined with sensors in the bowl. Petivity says the model was trained on more than 21,000 real-world feeding sessions, over 1,400 hours of video and images of more than 15,000 cats. The acceptance figure in the release, 99.5 percent, comes from a small test. The product page explains it as “199 out of 200 cats ate from the feeder in final testing”.

It works with Petivity’s Smart Litter Box Monitor, so owners who already use that device can “track consumption and elimination side by side” in the same app. A cat that eats less and visits the litter box less often shows up in both records at once.

The part it cannot do

Petivity’s product page includes a comparison table with three columns: its own feeder, standard automatic feeders and microchip feeders. In the row “Prevents food stealing”, the Petivity feeder gets a No and microchip feeders get a Yes. The page repeats it in plain words: “The feeder can’t stop one cat from eating another’s meal.”

What the feeder offers instead is a record. “If one cat’s meal is dispensed but another one eats it, you’ll find out,” the page says. That is useful if you are trying to find out why one cat is gaining weight. It does not help if one cat is on a prescription diet the other must not eat, because by the time the app tells you, the food is gone.

A microchip or RFID feeder keeps its lid closed until the right cat’s chip is read. So for strict diets, that design still wins. For households that only want to know who is eating how much, face recognition avoids a collar tag and needs no chip registration.

Specifications that matter before you order

The feeder holds 5 liters of kibble and dispenses between 1/8 cup, or 12 grams, and 3/4 cup, or 72 grams, per meal. It takes dry kibble from 5 to 15 mm across and “can’t support canned, wet food or raw food”. If any of your cats eats wet food, this feeder covers only part of its diet, and the intake figures in the app will be incomplete.

Power comes from a 5V 2A USB-C cord that is included. The backup is three D-cell alkaline batteries that are not included, and Petivity markets it as a 5-day backup battery. It measures 16.5 by 8 by 15 inches and weighs 10 lb. The app needs iOS 15 or later or Android 11 or later. Petivity offers a 90-day refund guarantee and a one-year limited warranty.

Petivity’s product page does not list the Wi-Fi band. It also does not say whether the camera records video, whether any footage or face data is stored in the cloud, or for how long. Those are fair questions for any camera placed in a kitchen, and our checklist for smart pet device security covers what to look for in a privacy policy.

How it compares on cost

A close rival is the PETLIBRO Granary 2 Vision Duo, a two-bowl camera feeder that recognizes which pet is at which bowl and weighs what each one eats with built-in scales. Live view and the scale data are free on that feeder, while recording, playback and eating-habit alerts sit behind PETLIBRO’s subscriptions.

Petivity has taken the opposite approach and put every feature of this feeder in the base price. Over three years, the difference is the full cost of whichever PETLIBRO plan you would otherwise buy. The trade is that the Granary 2 Vision Duo measures intake by weight at each bowl, while Petivity relies on a camera and bowl sensors at a single bowl, and Petivity has not published an accuracy figure for its recognition.

Who it suits

The Petivity Smart Feeder makes most sense in a home with two to five cats on the same dry food, where the question is who is eating what. It also fits owners who already use the Smart Litter Box Monitor and want food and litter data together.

It is the wrong choice for a cat on a prescription diet in a shared home, for wet or raw feeding, and for buyers outside the United States until Petivity names other markets. If you order, the 90-day refund window gives you time to see whether your cats accept it and whether the app’s per-cat records match what you see at the bowl.

That makes it a good fit for a home with one or more cats, or a small dog, where the goal is keeping other animals out and knowing from your phone when your own pets come and go. It is a poor fit if you need one pet kept inside while another roams, because this door cannot tell pets apart on the way out. Sure Petcare's smaller Microchip Cat Flap Connect can, and that is the model to buy for a cat-only household that needs per-pet rules.

The door has been on sale since 2017, it is the app-connected version of SureFlap's Microchip Pet Door, and on September 22, 2026 it was number 1 in Amazon's US Electronic Cat Doors best-seller list. The app features need the Sure Petcare Hub, which is sold separately. Nobody at PetTechScout has fitted or used this door. It rests on Sure Petcare's product and Hub pages, the full manual, the app store listing and retailer data, each cited where it is used.