BLOG

Testsy blog

Guides and engineering notes from the team building Testsy: how to check what your coding agent built, what we measured, and how the tester decides a verdict.

GuidesWebTesting an app you built with AI: a guide for foundersYou built it with Claude Code, Cursor, Lovable or Bolt and it looks done. How to check it works for a stranger and keeps working after every change.2 min read GuidesWebFind the bug before your client does: a guide for agenciesFor agencies and freelancers: check client sites before handoff, keep checking after launch, test forms to the inbox and checkout after updates, and show the proof.2 min read GuidesWebRelease testing for product teams: a guideCheck every deploy like a customer would without a QA department: which journeys to test, where to run them, how to hear about failures and how to keep the suite alive.2 min read GuidesiPhone and iPadAndroidMobile release testing: a guide for app teamsFor iOS and Android teams: tests that survive a redesign, App Store review, builds for CI, an agent that tests its own build, and the states other tools miss.2 min read GuidesWebYour coding agent said “done”. Here is what it did not checkAgents mark features complete after the code compiles and the unit tests pass. Why that misses broken screens, and how to check the feature as a user would.3 min read GuidesWebSign-up email links to localhost? Catch it before your users doA confirmation or reset email that points at localhost or an old preview is a common bug in AI-built apps. A test that opens the real email finds it on the first run.3 min read GuidesWebIt works in preview and breaks in production. Test the live addressAI app builders blur preview and production, so the version you clicked through is not the one users get. Test the deployed address after every deploy instead.2 min read GuidesWebFix one bug without breaking another: one bug, one testAI-built apps fall into a loop where each fix breaks something that worked. Keep a saved test for every bug you fix and run them all after each change.2 min read GuidesWebTurn a failed test into a fix prompt for your coding agentDescribing a bug to an AI agent is where fixes go wrong. Copy the fix prompt from a failed run instead: the check, what was seen, the steps and the proof.2 min read GuidesWebLet Cursor check its own work in a real browserConnect Testsy to Cursor over MCP so the agent can save a plain-language test, run it against your deployed app or a phone build, and read the verdict.3 min read GuidesWebLet Codex check its own work in a real browserAdd Testsy to Codex with one command. The agent drafts a plain-language test, runs it against your app, reads every check and reruns after its fix.2 min read GuidesWebTest from Gemini CLI, the Claude desktop app and claude.aiInstall Testsy as a Gemini CLI extension or a Claude connector, then ask in plain words for a browser or phone test and get a verdict with evidence.2 min read GuidesWebTest a Lovable, Bolt or Replit app by its address, no code neededBuilt your app in Lovable, Bolt, Replit or v0? Paste the published address, describe a user's journey in plain words, and get a verdict with a replay.2 min read GuidesWebSaved, then gone after a refresh: test that your app really savesA screen can show data that was never stored. One plain-language test creates a record, reloads the page and checks it is still there, the bug AI-built apps ship most.2 min read GuidesWebTest that one user cannot see another user's dataApps built with AI often show one account's records to another. A two-person test signs in as each, in separate browsers, and checks what each one can see.2 min read GuidesWebTest password reset from the email to the new passwordA reset email that links to the wrong host, or a link that fails in a new browser, locks users out. One test requests the reset, follows the link and signs in again.2 min read GuidesWebCatch the blank white screen after a deployThe most reported failure of AI-built apps is a page that renders nothing in production. A test that reads one value and records the console catches it on the first run.2 min read GuidesWebTen tests to run before you share your AI-built appA pre-launch checklist for apps built with Claude Code, Cursor, Lovable or Bolt, written as ten plain-language tests you can run on the live address today.2 min read GuidesWebWrite a test without writing code: steps, checks and valuesA plain-language test is only as good as its checks. How to write steps a tester can follow and checks it can judge, with examples that pass review first time.2 min read GuidesWebWhat an AI tester cannot check for youA browser test proves what a user can see and do. It does not audit security, load or code. An honest list of what Testsy covers, and what still needs a person.2 min read GuidesWebTest every Vercel preview before you mergeRun your saved plain-language tests against each pull request's preview, including previews behind Vercel Deployment Protection, and fail on a regression.2 min read GuidesWebTest Stripe checkout and plan upgrades in test modeBilling bugs cost money twice. A plain-language test goes through Stripe Checkout in test mode, comes back and checks the plan changed, plus declines and 3D Secure.2 min read GuidesWebRun browser tests after every deploy with GitHub ActionsA small set of plain-language release checks that run on each deployment, fail the job on a broken journey, and keep the replay next to the commit that broke it.2 min read GuidesWebTests that survive a redesign: why selector tests breakTest maintenance is the cost that kills end-to-end suites. Why tests tied to selectors fail after UI changes when the app still works, and what to write instead.2 min read GuidesWebCheck AI-written code before your customers doTeams ship more code than anyone reviews line by line. Checking each deploy from the outside, as a customer would, catches what review and unit tests miss.2 min read GuidesWebRun plain-language browser tests in GitLab CIInclude Testsy's GitLab template, point it at your project or a review app, and fail the pipeline on a broken journey, with JUnit results and watch links in the job.2 min read GuidesWebRun browser tests from any CI with one fileCircleCI, Bitbucket, Jenkins, Buildkite or Azure Pipelines: one dependency-free script runs your saved tests, writes JUnit and exits non-zero on a failure.2 min read GuidesWebRun your key journeys against production every nightDeploys are not the only thing that breaks an app. Email providers, payment settings and third-party scripts change too. A nightly schedule catches what no deploy caused.2 min read GuidesWebSend test failures to Slack, with the proof attachedConnect a Slack channel, subscribe it to the projects your team owns, and get failed, blocked and inconclusive runs with what was seen and a link to the replay.2 min read GuidesWebSend test results to Linear or Jira with webhooksTestsy posts each finished run to your endpoint as signed JSON. Verify the signature, then open an issue with the failed check, what was seen and the link to the proof.2 min read GuidesWebTest team invitations from the email to the joined workspaceInvitations cross two people, two sessions and an email. One plain-language test invites a second address, accepts from the email and checks both sides see the same team.2 min read GuidesWebTest admin and member roles: what each person can see and doPermission bugs show a member a button they should not have, or hide one an admin needs. Test each role with its own stored account and check the screen for both.1 min read GuidesWebRelease checks for a startup with no QA teamNo QA engineer, two deploys a day, and a manual click-through that gets skipped. A minimal release check that a small team can keep: five tests, one gate, one channel.2 min read GuidesWebFlaky tests, and what a test verdict should be allowed to sayA test that passes and fails on the same code teaches people to ignore it. Why a verdict needs room to say blocked or inconclusive, and how to find flaky tests.2 min read GuidesWebWhat a browser test suite costs when you pay per runNo seats and no minutes: one test run once is one run. How to estimate a month of release checks, schedules and CI runs on Testsy's plans, and what is never counted.2 min read GuidesWebTest search, filters and sorting without hard-coding the dataSearch boxes with suggestions, long lists, sorting and direct links are where list pages break. A plain-language test types, picks, scrolls, deletes and reads the result.1 min read GuidesWebTurn your release checklist into tests that run themselvesThe checklist in the wiki gets skipped on busy days. Move each line into a plain-language test, run them from CI or a coding agent, and keep the evidence per release.2 min read GuidesWebCheck the numbers: totals, prices and counts on screenA page that loads is not a page that is right. Make checks read the actual total, price or count, and keep the value the verdict was decided on as evidence.2 min read GuidesWebTest receipts, notifications and other emails your app sendsOrder receipts, mention alerts and weekly digests break without anyone noticing. A run with its own mailbox triggers the email, opens it and checks what it says.1 min read GuidesWebShare a live test run in a pull request, no account neededEvery run started from CI, the API or an agent comes with a watch link. Paste it in the pull request and reviewers watch the browser live, then read the result.2 min read GuidesWebA website launch QA checklist for agencies, with tests that runForms, checkout, redirects, phone layouts and email: the launch checks agencies list every time, written as plain-language tests to run before handoff and after.2 min read GuidesWebTest a contact form all the way to the inbox“Message sent” on the page proves nothing. How to test that a form's confirmation email arrives and the submission reaches the owner, with a run that has its own mailbox.2 min read GuidesWebCheck forms and checkout after every WordPress plugin updatePlugin, theme and core updates break forms and checkout while the site stays up. A scheduled plain-language test catches it the morning after, not weeks later.1 min read GuidesWebTest Shopify add to cart and checkout after an app or theme updateAn app install or theme update can silently break the cart. A plain-language test adds a variant, reaches checkout and pays with Shopify's test gateway, on a schedule.1 min read GuidesWebTest a booking form or calendar embed from start to confirmationBookings run through date pickers, embedded widgets from another company and confirmation emails. One plain-language test books a slot and checks every step.2 min read GuidesWebShow clients proof, not screenshotsA screenshot in an email proves one moment. A share link to a test result shows each check, the value it saw, a replay and a video, and expires when you choose.2 min read GuidesWebOrganise testing for many client sites: one project per clientHow agencies keep tests, sign-in accounts, schedules and run usage separate per client, when to use a project and when a separate workspace with its own billing.2 min read GuidesWebAdd functional checks to a website care planCare plans sell updates, backups and uptime. A nightly test that the form, booking or checkout still works is the check clients notice and understand.2 min read GuidesWebTest a site at phone, tablet and desktop sizesMost visitors arrive on a phone, and most QA happens on a laptop. Run the same plain-language test at 412 × 914, 820 × 1180 and 1280 × 800 and compare.2 min read GuidesWebTest form validation messages, not only the happy pathAn empty or mistyped form should say what is wrong, where. A plain-language test submits it empty, reads the exact error, then completes it and reads the saved result.1 min read GuidesWebTest redirects after a site redesign or migrationOld addresses from Google and other sites must land on the right new page. A plain-language test opens each old URL and checks where it ends up and what the page says.2 min read GuidesWebTest a newsletter signup with double opt-in, end to endA signup form that says “check your inbox” proves nothing until the confirmation email arrives and its link works. One test signs up, confirms and checks the welcome.1 min read GuidesWebTest a cookie banner: accept, decline, and the choice stickingRegulators expect rejecting cookies to be as easy as accepting them, and the choice to hold. A plain-language test clicks both, reloads, and checks the banner behaves.1 min read GuidesWebA handoff record that proves the site worked at launchScope arguments after launch start with “it never worked”. Run the key journeys at handoff and keep the results, with replays, as the record both sides can open.1 min read GuidesWebGive clients a viewer seat to see their test resultsA viewer sees projects, tests, runs and the proof, and changes nothing. When to invite a client as a viewer, when a share link is enough, and when to split workspaces.2 min read GuidesWebWhen bot protection blocks your tests, and how to allow themCloudflare-style challenges and firewalls stop automated browsers, including the tester. Why a blocked run is not counted, and how to allow Testsy's static addresses.2 min read GuidesWebTest a multilingual site and its language switcherThe switcher sends French visitors to an English page, or loses the page on switch. Test each language version, and browse from the country you need.1 min read GuidesiPhone and iPadPass App Store review: test what the reviewer will do firstApple says over 40% of unresolved review issues fall under guideline 2.1. Test the reviewer's path: install, sign in with the demo account, reach the content.2 min read GuidesiPhone and iPadAndroidTest an Expo app in CI without rebuilding it every timeNative builds are the slow part of mobile CI. Store an Expo fingerprint with each build, skip the build when it has not changed, and test the one you have.2 min read GuidesAndroidBuild an Android APK that runs on cloud test devicesWhy an APK needs both arm64-v8a and x86_64 native code to run on Android emulators, and the exact build command for React Native, Expo, Flutter and Gradle.2 min read GuidesiPhone and iPadBuild an iOS app for cloud testing: a zipped Simulator .appTestsy runs iPhone and iPad tests on an iOS Simulator build. The xcodebuild, Expo and Flutter commands that make one, how to zip it, and what to do in CI.2 min read GuidesiPhone and iPadAndroidTest a React Native app in plain English, without test IDsReact Native end-to-end suites are hard to set up and keep green. Write the journey in plain words, run it on iPhone and Android, and get a verdict with a replay.2 min read GuidesiPhone and iPadAndroidTest a Flutter app on iPhone and Android, canvas includedFlutter draws its own widgets, which trips up tools that read the native view tree. How Testsy tests Flutter apps, the iOS build to upload, and drawn content.2 min read GuidesiPhone and iPadAndroidTest mobile onboarding and the first launch, on a fresh installFirst launch is where new users decide. Each app run starts with fresh app data, so onboarding, prompts and the empty state are tested as a new user sees them.2 min read GuidesiPhone and iPadAndroidTest your app's iPad layout and landscape, not only the phoneApps built and tested on an iPhone break on an iPad or turned sideways. Run the same test on a tablet and in landscape, and check the layout-specific parts.2 min read GuidesiPhone and iPadAndroidOne test for iPhone and Android: write it once, run it on bothThe same plain-language test runs on an iOS build and an Android build. Where the two platforms differ, how to word a test for both, and what each result shows.2 min read GuidesiPhone and iPadAndroidTest a chat or marketplace app with two phones in one testMessages, offers, invitations and live updates need two people on two devices. One test can use two phones, and say what each person does and sees.2 min read GuidesiPhone and iPadAndroidLet your coding agent test the mobile app it just builtClaude Code, Cursor or Codex builds the app, uploads the build over MCP without an API key in the shell, runs your saved app tests and reads the verdict.2 min read GuidesiPhone and iPadAndroidMobile tests that survive a redesignMobile end-to-end suites fail when screens change and the app still works. Why locator-based tests break, and how a test written as behaviour keeps passing.2 min read GuidesiPhone and iPadAndroidSign in to a mobile app in a test without the password in the testMost app tests start behind a sign-in. Store the account once; the tester types the password into the app itself and the AI never sees it, on iPhone and Android.2 min read GuidesiPhone and iPadAndroidImport app builds from EAS, Bitrise or S3 by linkIf your pipeline already stores the APK or zipped .app somewhere, give Testsy the link. It fetches the build, checks it and runs the app's tests, with no upload step.1 min read GuidesiPhone and iPadAndroidTest a Capacitor or Ionic app: native shell, web contentCapacitor and Ionic apps are web pages inside a native shell. How the tester works in the web view, the build to upload, and why a debug build gives more detail.2 min read GuidesWebSign-up emails that stop arriving after launchSupabase's built-in mailer sends 2 emails an hour, only to your own team. Sign-up works for you and fails for everyone else. A test with its own mailbox catches it.2 min read GuidesWebPaid, but still on the free plan: test the Stripe webhookStripe takes the money, the app never hears about it. A test that pays in test mode and checks the plan changed catches it.2 min read GuidesWebUploads that work for you and fail for your usersA photo upload works when you try it and fails for a normal user with a 403 from storage rules. Test it as a normal account, with a real file, and check after a reload.2 min read GuidesWebTest your app as a brand-new user, not as yourselfYour account has data, a team and every flag on. A new user's has nothing. Sign up fresh on every run and check the first thing a new user must be able to create.2 min read GuidesWebTest that a double click does not create two ordersA slow Pay button gets clicked twice and the customer is charged twice. Test what a hurried person does, then check exactly one order exists, not the thank-you page.2 min read GuidesWebTest that a multi-step form keeps answers when users go backStep 3 of a checkout or sign-up, the user presses Back to fix step 1 and finds every answer gone. One plain-language test goes back and forth and checks each value.2 min read GuidesWebRun tests in production without polluting your dataTests against production leave accounts, orders and sign-ups in your numbers. Mark test data so it is easy to filter and clean, and keep the runs that only read separate.2 min read GuidesWebTest both sides of a feature flag before you flip itA flag doubles the app you ship: on and off. Bugs hide in the side nobody tested. Two stored accounts, one in the rollout and one out, run the same journey each.2 min read GuidesWebUpgrade a framework without breaking what users doA major Next.js or React upgrade builds, passes type checks and still changes behaviour, like caching. Run your saved user journeys against the upgrade's preview first.2 min read GuidesWebWrite tests in your team's own languageTests in Spanish, Vietnamese or German, with the plan and results in that language and the screen's text kept as it is. Two real runs, and the lesson one of them taught.3 min read GuidesWebCheck the record was saved, not just the success messageA thank-you message proves the page changed, not that anything was stored. Check the record on screen, or read your system through an endpoint you define.2 min read GuidesWebTest a WooCommerce checkout after every updateWooCommerce, payment plugins and the theme update on their own schedules, and the checkout is where they collide. A nightly test places an order and checks it arrived.2 min read GuidesWebAfter moving a site to a new host, check the forms still sendThe pages look identical after a migration, but contact forms and password resets often stop sending email. Test the flows that leave the site, not just the pages.2 min read GuidesWebTest what visitors in another country seePrices in the wrong currency, a missing EU cookie banner, a redirect to the wrong store. Tests can browse from the client's country, with its time zone and language.2 min read GuidesWebTest a lead magnet from the form to the inbox to the fileA free guide behind a form is three systems: the form, the email tool and the file. Any of them can break unseen. One test signs up, opens the email and checks the file.2 min read GuidesWebTest a membership site: members in, everyone else outCaching can show members' content to the public, or lock members out. One test signs in as a member and checks a visitor in a separate browser sees only the teaser.2 min read GuidesiPhone and iPadWhat survives when a user deletes and reinstalls your appOn iPhone, deleting an app clears its files but not its keychain, so a reinstall can come back signed in. A real run shows the note wiped and the token still there.3 min read GuidesiPhone and iPadFind unlabeled buttons and tiny tap targets in a mobile appApple's App Store now shows accessibility labels on product pages. An audit run on an iPhone, inside a normal test, catches low contrast, missing labels, small targets.2 min read GuidesAndroidTest the Android Back button after the Android 16 changeApps targeting Android 16 no longer get onBackPressed, and Google Play requires that target. A real run tests the Back button, a cancelled back gesture and split screen.2 min read GuidesAndroidThe keyboard covers the text field on Android 15 and 16Apps targeting Android 15 draw edge to edge and the window no longer resizes for the keyboard, so fields end up behind it. A real run checks the field stays visible.2 min read GuidesWebTest what users see when a request failsApps built with AI handle the happy path and show a blank page or a spinner forever when an API fails. Test the error message, the Retry button and the slow case.3 min read GuidesWebSign in with Google works locally and fails in production“Error 400: redirect_uri_mismatch” is what new users see when the Google sign-in was only set up for localhost. A test presses the button and checks where it lands.2 min read GuidesiPhone and iPadTest gestures in a mobile app: long press, swipe, drag and pinchGestures are where mobile UI tests are weakest and where gesture libraries clash. A real iPhone run does long press, swipes, drag to reorder, pinch and rotate.2 min read GuidesAndroidTest a long list on Android: scroll, swipe to delete, pull to refreshLists are where scrolling, swipe actions and pull to refresh compete for the same finger. A real Android run scrolls to item 75, swipes one away and refreshes.2 min read GuidesAndroid“Coupon applied” is not a discount: test the total, not the messageAn app says the coupon was applied and charges full price. A real Android run caught exactly that, because the check was on the subtotal, not on the message.2 min read GuidesiPhone and iPadTest photo and file pickers in a mobile appPicking a photo or a file runs through the phone's own screens, which app tests rarely reach. A real iPhone run picks a photo, a PDF, and a PDF where an image belongs.3 min read GuidesWebTest the phone menu on a client site after every updateThe hamburger menu is the navigation for most visitors, and it breaks after theme and plugin updates without anyone noticing on desktop. A phone-size test opens it.2 min read GuidesAndroidTest the files your mobile app exports, downloads and shares“Export complete” proves nothing about the file. A real Android run opens the CSV the app wrote, checks its rows, shares it through the share sheet and reads a download.2 min read GuidesWebThe hardest parts of web testing, each one run for realIframes, new tabs, downloads, uploads, drag and drop, slow pages, video, email codes, two sites at once. Why each is hard to automate, and a real run of each.5 min read GuidesWebTest a form inside an iframe, Shadow DOM and a new tabThree places a page hides its controls from a test script. One plain-language test fills a form in an iframe, clicks inside Shadow DOM and follows a link into a new tab.3 min read GuidesWebTest what is inside a download: CSV, Excel, PDF and ZIPA file that arrives is not a file that is right. Two plain-language tests download five files, open each one and compare what is inside with what the page shows.3 min read GuidesWebTest file uploads: a good file, a fake one and one too bigUpload tests need a folder of awkward files nobody keeps. Two plain-language tests upload a photo, a text file named .jpg, a 52 MB image, two CSVs and two documents.3 min read GuidesWebTest drag and drop, resizing and drawing on a canvasDragging is a gesture, not a click, and a canvas has nothing inside it to select. One plain-language test drags, reorders, resizes and draws, and reads each result.3 min read GuidesWebTest alert, confirm and prompt, and the browser's permission promptsThese boxes belong to the browser, not the page, and they stop everything until answered. One plain-language test accepts, cancels, types an answer, allows and blocks.2 min read GuidesWebTest that sign-in and saved data survive a reloadThe page looked right until someone pressed refresh. One plain-language test signs in, reloads, saves a note, reloads again and checks what is left.3 min read GuidesWebTest slow pages and failed requests without a single sleepA button that is not ready, a report still loading, a request that fails, a page that never finishes. Three plain-language tests with no wait times in them.3 min read GuidesWebOne web test across a frame, an upload, a second tab and a downloadReal workflows cross every boundary a script hates in one sitting. One test uploads a file in a sandboxed frame, approves in another tab and reads the manifest.3 min read GuidesWebTest that a video really plays, and what the page says when it cannotA player on the page is not a video that plays. One plain-language test presses play, checks the time moves, checks captions, plays audio and finds a broken video.3 min read GuidesWebTest date pickers, hover, right-click menus and keyboard shortcutsCustom widgets are where plain form tests stop. Two plain-language tests sort a table, pick dates two ways, format text, hover, right-click and use the keyboard.4 min read GuidesWebTest email verification codes, the wrong code and the resend tooA sign-up test that stops at "we sent you a code" proves little. Two real runs read the email, try a wrong code, ask for a new one and finish signing up.5 min read GuidesWebTest an admin site and a public site, as two people, in one testWhat an editor publishes, a visitor must see, and a draft must stay hidden. Two real runs cross from an admin site to a public one, in two separate browsers.5 min read GuidesiPhone and iPadAndroidThe hardest parts of mobile app testing, each one run for realPermission prompts, notifications, SMS codes, Face ID, location, crashes, old OS versions. Why each is hard to test, and a real run of each with its result and video.5 min read GuidesiPhone and iPadAndroidTest that a notification opens the right screenThe notification arrived. Did tapping it open the order, or only the home screen? A plain-language test waits for it, taps it and checks where the app lands.3 min read GuidesiPhone and iPadTest what your app does when the user says no to a permissionLocation allowed, camera denied, microphone denied, contacts allowed. One plain-language test answers each system prompt as told and checks what the app shows after.3 min read GuidesAndroidTest SMS code sign-in on Android without a second phoneA text message arrives, Android offers the code, the app verifies it. One plain-language test covers the whole hand-off, with the message and the result on video.3 min read GuidesiPhone and iPadTest deep links: does the link open the right screen?A link in an email or an ad should land on the item, not the home screen. A plain-language test opens the link on the phone and checks the screen the app shows.3 min read GuidesiPhone and iPadTest Face ID unlock: the wrong face, then the right oneThe locked path matters as much as the unlocked one. A plain-language test answers the Face ID prompt with the wrong face, then the right one, and checks both.2 min read GuidesiPhone and iPadTest what your app remembers after the phone kills itPhones close apps in the background to free memory. A plain-language test types a draft, has the phone kill the app, reopens it and checks the draft is still there.3 min read GuidesiPhone and iPadA mobile app crash, caught with the file and line that caused itWe asked for an order confirmation. The app crashed instead. The result names the error, the step it came after, the stack trace and the steps to reproduce it.3 min read GuidesAndroidTest your app offline, on low battery and during a phone callA phone loses signal, runs low and rings. Two plain-language tests on Android switch the network off and on, set the battery to 15% and take an incoming call.3 min read GuidesiPhone and iPadTest rotation, dark mode and the largest text sizeThree phone settings change every screen in your app at once. Two plain-language tests flip each one mid-run and check that the app still reads and still works.3 min read GuidesiPhone and iPadTest location features without leaving your deskSet where the phone is, move it along a route at walking or driving speed, and check what your app shows. Two real runs, and one passing check that hid a bug.3 min read GuidesiPhone and iPadTest a Live Activity and a home-screen widget on iPhoneThey are your app, shown where your app is not: the lock screen, the Dynamic Island, the home screen. One plain-language test starts one, adds the other and reads both.3 min read GuidesiPhone and iPadAndroidOne mobile test, six OS versions: what changed on screenWe ran the same plain-language tests on iOS 17, 18 and 27 and on Android 10, 12 and 16. The prompts differed on every version. The test text never changed.5 min read MigrationMoving from TesterArmy to Testsy before 15 JanuaryTesterArmy's dashboard closes after 15 January 2027. Here is what carries over to Testsy, what does not, and how to move a plain-English test suite.5 min read EngineeringWebMaking an AI tester prove its verdictsAn agent that drives a browser can say "passed" about almost anything. These are the rules we built Testsy on so that every verdict comes with its evidence.5 min read GuidesWebMake Claude Code test what it just builtYour coding agent says the feature works. Connect Testsy over MCP and it has to run the feature in a real browser and bring back proof before it says done.4 min read GuidesWebTest sign-ups that send an email or SMS codeA sign-up form that submits is not a sign-up that works. Give the test its own mailbox or phone number and check the whole journey, code included.4 min read GuidesWebA daily check on every client site's key flowClients find out their booking form is broken from their customers. Set up one scheduled test per site, get told when it fails, and send the client the proof.4 min read ComparisonsWebFive AI testers, one planted bug: what each verdict rested onWe gave the same tests to five AI testing tools and to our own, one with a bug planted in a download. One tool passed it and one excused it. Here is what we saw.7 min read