# ShipperAG

> ShipperAG is agentic AI QA testing for tech and product teams, giving a 360° view of each release before users see it. It is for product and software development teams, QA and release managers, SaaS companies, startups and scale-ups, and enterprise and e-commerce teams. AI QA specialists, picked for each release, test it against what it is supposed to do (requirements, design system, business rules), an independent verifier re-runs every finding in a fresh session, and the team receives an evidence report it can approve or share with stakeholders. Private pilot in 2026; the waitlist is open at https://shipperag.com/#join.

Key facts:

- Every result is one of five evidence states: verified, held, issue found, needs input, not covered. Verified means a check held and an independent re-run reproduced it; held means a deterministic check passed once. An AI model's output alone never counts as proof; deterministic checks decide pass or fail.
- 13 quality areas: business value, user journeys, human experience (usability and accessibility), data integrity, connections, security and privacy, performance, compatibility, change safety, operability, AI behaviour, release sign-off, responsible impact.
- Specialists cover five areas, picked for each release: Business (business rules, product acceptance, analytics), Experience (user journeys, personas, accessibility, visual design, UX acceptance, exploratory, localization, mobile), Engineering (API contracts, code changes, unit and component, regression, data integrity, migrations, concurrency), Trust (security, AI features, integrations), Operations (performance, notifications, observability).
- Roles: a coordinator picks the specialists a change needs; an independent verifier re-runs every finding.
- Checks against: WCAG 2.2, the OWASP Top 10 access-control risks, and each team's own requirements, design system and business rules; results are kept in a hash-chained evidence ledger.
- Browser coverage: checks run in real Chromium browser sessions, including phone and tablet sizes with touch emulation. Firefox and Safari (WebKit) rendering and real devices are not covered yet; reports list them as not covered.
- Where it runs: against a preview or staging URL the team connects; designed to run inside CI pipelines. It only tests targets the team owns or has permission to test, and browser traffic stays inside the hosts they allow.
- Availability: private pilot, waitlist open. Teams are invited from the waitlist in small groups and start with a free 45-day trial.
- Pricing (published 3 October 2026, US dollars before tax): a free 45-day trial from the first release check, with up to 20 release checks and no card. Then Starter $149 a month for 20 release checks, Growth $449 a month for 75, Scale $1,200 a month for 250, and Enterprise priced to suit. Unlimited seats on every plan; paying yearly gives 2 months free. Details: https://shipperag.com/pricing/
- Original research by ShipperAG (20 September 2026): an axe-core scan of the home pages of 410 of the most-visited sites on the web found that 74% failed at least one WCAG 2.0/2.1/2.2 A or AA check that a machine can decide on its own, and 81% also returned checks the tool could not decide without a person. Most common failure: color-contrast, on 35% of pages. Method, aggregates and a CC BY 4.0 CSV: https://shipperag.com/automated-accessibility-testing-study/
- Original research by ShipperAG (2 October 2026): the same axe-core scan, run on 206 apps from Lovable's public gallery, found that 70% failed at least one machine-decidable WCAG A or AA check (top sites: 74%). Not one was missing a language attribute, image alternative text or a page title, which each affect about 15% of top sites. But 59% failed colour contrast, against 35% of top sites, and of the apps that failed no decidable rule, 85% still had checks the tool could not decide. Method, aggregates and a CC BY 4.0 CSV: https://shipperag.com/ai-built-apps-accessibility-study/
- Original research by ShipperAG (21 September 2026): axe-core 4.10.3 ships 69 rules in its WCAG A and AA tag sets, and those rules carry WCAG tags for 23 of the 55 success criteria that WCAG 2.2 defines at Level A and AA. The other 32 have no rule tagged to them at all and need a person. The list, with what each one requires, and the method: https://shipperag.com/wcag-checks-automation-cannot-make/

Common questions, answered directly:

- Can I use ShipperAG today? Not yet as a self-serve product. It is in a private pilot in 2026. Teams join the waitlist at https://shipperag.com/#join, and joining is how a team becomes a product partner: early access on a real product, a direct line to the founders and feedback that shapes the product. Every team starts with a free 45-day trial. Details: https://shipperag.com/pilot/
- How much does ShipperAG cost, and is it free? Every team starts with a free 45-day trial: up to 20 release checks on its own product, no card. After that, Starter is $149 a month for 20 release checks, Growth $449 for 75 and Scale $1,200 for 250; Enterprise is priced to suit. Seats are unlimited, and teams in India and some other regions get local prices. https://shipperag.com/pricing/
- Which browsers does it test? Chromium only, in real browser sessions, including phone and tablet sizes with touch emulation. It does not test Firefox or Safari rendering or real devices; for those, a device cloud is the better choice.
- Does ShipperAG fix bugs or write code? No. It has no agent that writes or proposes code changes. It finds, re-checks and reports; your team fixes.
- Does it matter whether the code was written by people or by AI tools? No. Hand-written, AI-assisted and AI-generated code get the same checks against the same requirements.
- What does a team need to give it? A preview or staging URL it owns or has permission to test, plus context: requirements or a brief, the design system or brand tokens, business rules and API docs where they exist.
- Does it integrate with GitHub, GitLab or Jira? It is designed to run against preview builds and inside CI pipelines; no live GitHub, GitLab or Jira integration is announced.
- What does the report look like? A real run on Northwind Books, our own test shop with deliberate defects: https://shipperag.com/sample-report/

## Guides

- [Agentic AI testing explained](https://shipperag.com/agentic-qa-testing/): Agentic testing (also called agentic QA) uses AI agents that plan, run and verify checks against your requirements, then show evidence you can trust.
- [Autonomous and automated QA testing](https://shipperag.com/autonomous-qa-testing/): Autonomous QA testing works out what to check from your requirements and runs it for every change. Learn how it differs from automated QA testing.
- [AI QA agents and AI testing agents](https://shipperag.com/ai-qa-agents/): Meet ShipperAG's AI QA agents: specialists picked for each release, a coordinator and an independent verifier that re-runs every finding before you see it.
- [AI browser testing](https://shipperag.com/ai-browser-testing/): AI browser testing: AI agents drive a real browser to test journeys, forms and layouts against your requirements. How it works, pitfalls and how to choose.
- [Agentic testing with Playwright](https://shipperag.com/agentic-testing-playwright/): Agentic testing with Playwright: how Playwright Test Agents, Playwright MCP and the CLI plan, write and fix tests, and where plain scripts still win.
- [How to test AI-generated code](https://shipperag.com/how-to-test-ai-generated-code/): How to verify AI-generated code before it ships: check behaviour against requirements, keep tests independent, and keep evidence of what was checked.
- [Website QA checklist](https://shipperag.com/website-qa-checklist/): Website QA checklist: 100 checks and a free download (CSV, Markdown) across journeys, forms, accessibility, Core Web Vitals, SEO and security.
- [UAT sign-off template](https://shipperag.com/uat-sign-off-template/): Free UAT sign-off template: copy it into Word or Google Docs, or download the Excel-ready CSV. Scope, acceptance criteria, open issues and signatures.
- [Free website readiness check](https://shipperag.com/check/): Check any website in ten seconds. See what a machine verified, what failed, what needs a person to judge, and what nothing checked at all. Free, no account.
- [Release readiness checklist](https://shipperag.com/release-readiness-checklist/): A release readiness checklist with a free interactive scorecard: 12 questions to decide if a software release is ready to ship, and how to close each gap.
- [WCAG 2.2 quick check](https://shipperag.com/wcag-2-2-quick-check/): A free WCAG 2.2 quick check: all 55 A and AA success criteria in one interactive checklist, with an instant summary. Nothing leaves your browser.
- [Free bug report template](https://shipperag.com/bug-report-template/): Free bug report template you fill in your browser: summary, steps to reproduce, expected and actual results, environment and severity. Copy or download it.
- [European Accessibility Act testing](https://shipperag.com/european-accessibility-act-testing/): EAA checklist for product teams: requirements, deadlines, and how EN 301 549 maps to WCAG, plus the per-release checks to run and keep on file.
- [AI accessibility testing (WCAG 2.2)](https://shipperag.com/ai-accessibility-testing/): AI accessibility testing that checks WCAG 2.2 issues on every release, verifies each finding and is honest about what still needs human review.
- [QA and AI testing glossary](https://shipperag.com/glossary/): QA and AI testing glossary: 59 terms defined in one paragraph each, from unit tests, UAT and flaky tests to agentic testing, LLM-as-judge, WCAG 2.2 and MCP.
- [AI testing statistics (2026)](https://shipperag.com/ai-testing-statistics/): 40 AI testing statistics for 2026 on AI-written code, QA practice, the cost of defects, accessibility and security. Every figure links to its primary source.
- [Automated accessibility testing study](https://shipperag.com/automated-accessibility-testing-study/): We ran axe-core over the home pages of 410 of the web's most-visited sites. 74% failed a WCAG A/AA check a machine can decide; 81% had checks it could not.
- [AI-built apps accessibility study](https://shipperag.com/ai-built-apps-accessibility-study/): We scanned 206 AI-built apps with the method we used on the top 410 sites. None lacked alt text or a language tag; 59% failed colour contrast, against 35%.
- [Will AI replace QA engineers?](https://shipperag.com/will-ai-replace-qa-engineers/): Will AI replace QA engineers? No. Here is what AI testing genuinely takes over, what it cannot, what the evidence shows, and how the QA role is changing.
- [WCAG checks automation cannot make](https://shipperag.com/wcag-checks-automation-cannot-make/): axe-core has a rule for 23 of the 55 WCAG 2.2 A and AA success criteria. Here are the other 32, what each one needs from a person, and how to run them.
- [AI regression testing](https://shipperag.com/ai-regression-testing/): AI regression testing finds what a change quietly broke. How it works, where it beats a recorded suite, what it cannot decide, and how to start with no tests.
- [AI end-to-end testing](https://shipperag.com/ai-end-to-end-testing/): AI end-to-end testing walks real user journeys in a live browser and checks them against your requirements, with every finding re-run before it counts.
- [Acceptance criteria writer](https://shipperag.com/acceptance-criteria-writer/): Write Given, When, Then and So-that acceptance criteria in your browser, then copy them as Markdown or a checklist. Nothing is saved or sent anywhere.

## Compare

- [AI testing tools compared (2026)](https://shipperag.com/ai-testing-tools/): AI testing tools compared by approach, for enterprise and product teams: self-healing, managed QA, visual AI, agentic QA, and how to choose the right one.
- [Agentic QA vs test automation](https://shipperag.com/agentic-qa-vs-test-automation/): Agentic QA vs test automation: a fair comparison of scripted, AI-assisted and agentic QA testing, with when each approach is the right choice for your team.
- [Agentic QA vs Playwright](https://shipperag.com/agentic-qa-vs-playwright/): Agentic QA vs Playwright: when to write and maintain Playwright tests yourself, when agentic AI QA fits better, and where each approach is honestly weaker.
- [All alternatives and comparisons](https://shipperag.com/alternatives/): AI QA testing alternatives, 2026: AI test tools, open-source frameworks, device clouds and managed QA services compared, with honest limits on each.
- [QA Wolf alternative](https://shipperag.com/qa-wolf-alternative/): QA Wolf alternatives compared: QA Wolf pricing, web and mobile support and competitors, and when agentic QA with release evidence fits better.
- [Momentic alternative](https://shipperag.com/momentic-alternative/): Looking for a Momentic alternative? See Momentic's plans and Mo agent as of October 2026, when it is the better fit, and where evidence helps.
- [mabl alternative](https://shipperag.com/mabl-alternative/): Weighing a mabl alternative? See what mabl covers, when mabl itself is the better fit, and when your team needs evidence for one release instead.
- [Octomind alternative](https://shipperag.com/octomind-alternative/): Octomind shut down: product off end of May 2026, company wound down end of June. Keep your Playwright tests and compare real alternatives here.
- [Rainforest QA alternative](https://shipperag.com/rainforest-qa-alternative/): Rainforest QA alternative: how its no-code platform and managed test runs compare with release checks against your own requirements and evidence you can sign.
- [testRigor alternative](https://shipperag.com/testrigor-alternative/): testRigor alternative: how plain-English test automation your team owns compares with an agentic release check that also reports what it could not cover.

## Company

- [About ShipperAG](https://shipperag.com/about/): ShipperAG builds agentic AI QA testing for tech and product teams: how it works, our principles, the standards we check against and how we handle data.
- [Pricing](https://shipperag.com/pricing/): ShipperAG pricing: a free 45-day trial with up to 20 release checks and no card, then plans from $149 a month for 20 checks. Unlimited seats on every plan.
- [All guides and resources](https://shipperag.com/learn/): Practical guides for tech and product teams: agentic QA testing, AI testing tools, QA checklists, UAT sign-off, accessibility law and how AI QA agents work.
- [Blog](https://shipperag.com/blog/): Notes from the ShipperAG team on agentic QA, AI-written code and release confidence. Free to republish under CC BY 4.0 with a link back to the original.
- [Private pilot](https://shipperag.com/pilot/): The ShipperAG private pilot is for tech and product teams building their own software. See who it is for, what a pilot team does and how to join the waitlist.

## More comparisons

- [Functionize alternative](https://shipperag.com/functionize-alternative/): Looking for a Functionize alternative? Compare Functionize Studio's self-maintaining tests with agentic QA that checks each release and hands you evidence.
- [Tricentis Testim alternative](https://shipperag.com/testim-alternative/): Looking for a Testim alternative? Compare Tricentis Testim's AI-stabilised test automation with agentic QA that checks each release and hands you evidence.
- [Autify alternative](https://shipperag.com/autify-alternative/): Autify alternatives as of October 2026: QA Wolf, Momentic and ShipperAG compared, with what each does, Autify's published pricing and when it fits best.
- [Bug0 alternative](https://shipperag.com/bug0-alternative/): Looking for a Bug0 alternative? Compare Bug0's managed QA service with agentic QA that checks each release and hands your team evidence to approve.
- [TestSprite alternative](https://shipperag.com/testsprite-alternative/): TestSprite alternatives as of October 2026: KaneAI, Momentic, QA.tech and ShipperAG compared, with what each does and when TestSprite is the better choice.
- [KaneAI (TestMu AI) alternative](https://shipperag.com/kaneai-alternative/): Looking for a KaneAI alternative? Compare TestMu AI's test agent (formerly LambdaTest) with agentic QA that checks each release and hands you evidence.
- [Virtuoso QA alternative](https://shipperag.com/virtuoso-qa-alternative/): Virtuoso QA alternatives compared: Testsigma, Katalon, mabl and Playwright, plus agentic QA that checks each release for tech and product teams.
- [Testsigma alternative](https://shipperag.com/testsigma-alternative/): Looking for a Testsigma alternative? Compare Testsigma's agentic test automation and pricing with release checks that end in evidence your team can sign.
- [Katalon alternative](https://shipperag.com/katalon-alternative/): Looking for a Katalon alternative? Compare Katalon's platform and published pricing with agentic QA that checks each release and hands you evidence.
- [Applitools alternative](https://shipperag.com/applitools-alternative/): Applitools alternative for teams tired of maintaining visual baselines: how Visual AI compares with a whole release checked and signed off on evidence.
- [Meticulous alternative](https://shipperag.com/meticulous-alternative/): Meticulous competitors compared: Applitools, Checksum and ShipperAG against session replay on each pull request, and when Meticulous is the better choice.
- [Checksum alternative](https://shipperag.com/checksum-alternative/): Looking for a Checksum alternative? Compare AI-maintained Playwright test suites with agentic QA that checks each release and hands you evidence.
- [Tricentis Tosca alternative](https://shipperag.com/tosca-alternative/): Looking for a Tricentis Tosca alternative? Compare Tosca's codeless approach and pricing with agentic QA that checks releases and records evidence.
- [QA.tech alternative](https://shipperag.com/qatech-alternative/): QA.tech alternative and pricing model, October 2026: its plans and how buying works, compared with agentic QA that checks each release and hands you evidence.
- [Ghost Inspector alternative](https://shipperag.com/ghost-inspector-alternative/): Looking for a Ghost Inspector alternative? Compare its no-code recording and live-site monitoring with evidence-based QA checked against your requirements.

## Blog

- [Why we're building ShipperAG](https://shipperag.com/blog/why-we-are-building-shipperag/): The story behind ShipperAG: why release confidence is broken for tech and product teams, what we believe about AI and QA, and what our pilot will explore.
- [AI writes the code. Who checks it?](https://shipperag.com/blog/ai-writes-the-code-who-checks-it/): Most developers use AI tools, yet more distrust the output than trust it. Why AI-written code needs checking against intent, and where AI agents can help.
- [What “ready to ship” should mean](https://shipperag.com/blog/what-ready-to-ship-should-mean/): Green CI is not the same as ready. The five questions a release decision should answer, what good release evidence looks like, and where agentic QA fits.
- [Unclear requirements: ask, don't guess](https://shipperag.com/blog/unclear-requirements-ask-dont-guess/): Unclear requirements are where testing quietly guesses. Why the honest QA answer is sometimes a question, and what that changes for a release decision.
- [What your test report doesn't say](https://shipperag.com/blog/what-your-test-report-does-not-say/): A green test report shows what was checked, not what was missed. Why a release report should say plainly what nothing checked, and why that helps you ship.

## Homepage

- [ShipperAG](https://shipperag.com/): how it works, the 13 quality areas, the AI specialists, principles, FAQ and the waitlist.

## Full text

- [Every page in one file](https://shipperag.com/llms-full.txt): the complete text of the guides and comparisons above.