Vibe testing
Vibe testing is the QA workflow where you paste a URL plus one sentence and a goal-based test runs against the deployed product.
Vibe testing is the discipline of producing browser evidence for the thing you just shipped — without writing a test suite first. You paste a URL and a single sentence describing the goal; an AI planner authors the scenario; Playwright (including mobile web) executes it; a vision model reviews the screenshots; you walk away with a token-gated public proof page. Vibe testing is the QA-side counterpart to vibe coding — the same ship-fast, AI-native philosophy applied to verification. Not mood or culture assessment, not water-well testing.
Read the full definition →
Vibe testing platform
A vibe testing platform runs web (Playwright, including mobile web) on the same AI planner, the same /app shell, and the same release-health timeline.
A vibe testing platform is the multi-engine counterpart to a single-purpose tool. Real browser runs for web (Chromium, Firefox, WebKit, mobile web) share one planner, one shell, and one timeline today (Chromium on every plan; Firefox, WebKit and mobile web on Team and Scale). A solo founder signs in once, picks a saved scenario from their phone at /app/vibe, hits Run, and gets a shareable proof page. The platform category is distinct from test management, CI orchestration, or QA-as-a-service.
See the platform category →
Vibe testing SaaS
A vibe testing SaaS delivers vibe testing as a cloud subscription — pricing tiers, no on-prem install, and the execution plane runs on the vendor's servers.
A vibe testing SaaS is the cloud-delivery counterpart to running your own QA infrastructure. You pay a monthly or annual fee, the execution plane runs on the vendor's infrastructure, and you open /app/vibe on your phone to run browser tests. Builder $29/mo, Team $99/mo, Scale $299/mo plans segment by browser access, monthly credits, triggers and capacity. Only passing runs use credits. Enterprise adds negotiated capacity.
See the SaaS pricing tiers →
AI vibe testing platform
An AI vibe testing platform uses an AI planner to author test scenarios from a URL plus a one-sentence goal, then dispatches a real engine to execute.
An AI vibe testing platform is the category Well Tested files under. The AI planner is responsible for authoring steps, surfacing assumptions and missing context, and reporting confidence. The real engine is responsible for execution. The vision model is responsible for scoring the screenshots against the goal. A solo founder does not write a multi-step agent prompt or a Playwright spec — they paste a URL and one sentence, then review the planner's assumptions before run.
See the multi-engine breakdown →
Multi-engine vibe testing
Multi-engine vibe testing runs web flows (Playwright, including mobile web) from one planner, one shell, and one release-health timeline.
Multi-engine vibe testing is the practice of routing a single AI-authored scenario to the engine that matches the goal — Playwright for browser flows (including mobile web) — and writing the evidence into the same release-health history.
Vibe testing tool
A vibe testing tool is a single-purpose product that helps you run, record, or share a vibe test. The category spans planners, runners, and proof-page generators.
A vibe testing tool sits inside the vibe testing workflow at one stage: the planner (AI authoring the steps), the runner (Playwright executing), or the proof-page generator (the shareable evidence). Well Tested ships planner, runner, and proof on the same shell — and is therefore a vibe testing platform, not a vibe testing tool.
See vibe testing vs vibe testing tools →
Vibe testing in QA
Vibe testing in QA is the practice of running a goal-based, AI-authored test against the deployed product and capturing the evidence as a release artifact.
Vibe testing in QA means treating the proof page as a first-class release artifact — alongside the commit, the deploy URL, and the changelog entry. The QA lead opens the share link in Slack, sees the screenshots and the vision verdict, and either signs off or escalates. Vibe testing is not a replacement for unit tests, integration tests, or code review — it is the verification layer above all of them, where the goal is "did the user actually experience what we shipped?"
Read the QA definition →
Playwright
Playwright is the open-source browser-automation library used by Well Tested to run web vibe tests in Chromium, Firefox, and WebKit.
Playwright is the engine behind every web vibe test on Well Tested. It supports headless and headful Chromium, Firefox, and WebKit, plus iOS Safari and Android Chrome via mobile web emulation. Multi-browser parallel runs are available on Team and Scale; Builder ships Chromium only. Playwright MCP is the Model Context Protocol server that wraps Playwright as a tool for AI agents — see Playwright MCP below.
Playwright MCP
Playwright MCP is the Model Context Protocol server that exposes a Playwright browser session as a tool for AI agents to drive.
Playwright MCP wraps Playwright in the Model Context Protocol — the AI-agent tool-integration layer. It gives an AI agent a browser handle and expects the agent to author the prompt loop, dispatch the browser steps, and parse the result. Well Tested uses Playwright MCP internally as one of its execution layers, but adds an AI planner (so the customer does not author the agent prompt), a vision verdict (so the customer does not parse the screenshots), and a shareable proof page (so the customer does not write a session log). Well Tested is therefore a Playwright MCP alternative for builders who want the outcome, not the browser handle.
Playwright MCP alternative →
Mobile testing shell
The mobile testing shell is the /app web app at welltested.io/app — pick a saved scenario, hit Run, watch status, share the proof link — from your phone.
The mobile testing shell is the customer-facing UI for running vibe tests from a phone. You sign in at /app/vibe, pick a saved browser scenario from a list, hit Run, watch status push back from the worker, and tap to copy a token-gated public link. The shell is a hosted web app — no app store install. Available on every plan.
See the mobile shell in context →
Run checks from your phone
Run checks from your phone means opening /app/vibe, picking a saved browser scenario, and watching status push back as the run completes.
Run checks from your phone is for solo founders and on-call engineers who want browser proof without a laptop. Open /app/vibe on your phone for sign-in, scenario selection, run dispatch, status polling, and proof-page sharing. Available on every plan.
Playwright MCP alternative
A Playwright MCP alternative is a product that gives an AI agent a browser session without requiring the customer to write the prompt loop, dispatch the steps, or parse the screenshots.
A Playwright MCP alternative is for builders who want the outcome of an AI-driven browser run — a vision verdict on the screenshots, a shareable proof page, a release-health entry — without writing a multi-step agent prompt. Well Tested wraps Playwright MCP internally and adds the planner, the vision verdict, the proof page, and /app/vibe run-from-phone on top.
Playwright MCP alternative comparison →
AI planner
The AI planner is the LLM that turns a URL plus one sentence into a structured test plan with assumptions, missing context, browser steps, and a confidence score.
The AI planner is the first stage of every Well Tested vibe test. It receives a URL and a one-sentence goal; it returns a structured plan with the steps it intends to run, the assumptions it is making about the page's intent, the context it wishes the customer had given it, and a confidence score from 0 to 100. The customer can push back on the plan before run. The planner is not the runner — Playwright is. The planner is the thing that decides what to run.
Vision verdict
The vision verdict is the LLM's review of the run screenshots — did the evidence actually prove the goal stated in the URL plus one sentence?
The vision verdict is the last stage of every Well Tested vibe test. The vision model receives the step screenshots, the planner's goal, and the planner's intended assertions; it returns a pass/fail verdict with a one-paragraph explanation and a per-step confidence breakdown. The verdict is what makes the proof page useful — it separates what the engine observed from what the planner inferred, and it surfaces exactly which step failed when something goes wrong.
Shareable proof page
A shareable proof page is a token-gated public URL that shows the screenshots, the vision verdict, and the run metadata for one Well Tested vibe test.
A shareable proof page is what you get when a vibe test finishes. It is a public URL at /vibe/<slug>/share?t=<token>; the token gate prevents SEO indexing and lets the customer revoke access. The page shows the planner's goal, the steps the engine ran, the screenshots at each step, the vision verdict, and a copy-pasteable badge for the marketing site. Shareable proof is the single most important output of the vibe testing loop — it converts a private run into a release artifact.
Release-health dashboard
The release-health dashboard is the per-app history of vibe runs — trends, recurring bugs, regressions, and the vision verdict that explains each one.
The release-health dashboard is the long-term value layer on top of the per-run proof page. Every run feeds a release-health entry: planner goal, engine, verdict, screenshots, step counts, and a link to the full proof. Trends emerge over time — a button that failed on every deploy, a checkout flow that regressed after the new pricing page, a step latency that crept up over three weeks. Available on Team and Scale.
Real engine
A real engine is the Playwright worker that actually executes the test — distinct from a simulated, mocked, or stubbed response.
A real engine in the vibe testing context means Playwright actually driving a Chromium browser and measuring what a real user would have experienced. Vibe testing insists on real engines because the entire point of the loop is to capture what a real user would have experienced.
Token-gated share
A token-gated share is a public URL with a private token — anyone with the link can view the proof, but the URL is not indexed by crawlers and the customer can revoke it.
Token-gated share is the privacy primitive under the shareable proof page. The share page lives at /vibe/<slug>/share?t=<token>; without a valid token the page returns 404 to crawlers and 401 to humans. The token is per-share and can be rotated from the customer dashboard. The page is intentionally not indexable by Google or AI crawlers — only the customer who runs the test decides who sees the proof.
On-the-move mobile shell
The on-the-move mobile shell is the /app web app — pick a saved scenario, hit Run, watch status, share the proof link — from a phone in a coffee shop.
The on-the-move mobile shell is the same /app shell as the mobile testing shell, with an emphasis on the customer being away from their laptop — on the train, in a meeting, on the couch. The shell is responsive, push-status-aware, and tap-friendly. It is the proof that vibe testing is not just browser automation — it is the workflow you can complete from your phone.
Goal-based test
A goal-based test is a vibe test specified by a URL and a one-sentence goal — the planner infers the steps, the engine executes them, the vision model scores the result.
A goal-based test is the input shape vibe testing was built around. The customer does not write a multi-step test plan or a Playwright spec. The customer states the goal — "the user can complete checkout with a saved card" — and the planner, the engine, and the vision model do the rest. This is the opposite of script-based testing, where the customer authors every step.
Browser proof
Browser proof is the per-step screenshot sequence captured by the Playwright engine during a web vibe test — the visual evidence the vision model reviews.
Browser proof is one half of the evidence a vibe test ships with (the other half is the vision verdict on those screenshots). Every step in the planner's plan gets a screenshot at the moment the engine considers the step complete; the sequence of screenshots is what the vision model reviews to decide pass/fail. Browser proof is stored on the shareable proof page and on the release-health entry for the run.
Pricing tier
A pricing tier is one of the four Well Tested plans (Builder, Team, Scale or Enterprise) that segments browser access, monthly credits, triggers and capacity.
Well Tested pricing tiers are: Builder $29/mo (300 credits), Team $99/mo (1,000 credits), Scale $299/mo (3,000 credits), and Enterprise (custom, negotiated capacity). Builder runs Chromium; Team and Scale add Firefox, WebKit, iPhone and Android mobile-web profiles and the release-health dashboard. Run-from-phone at /app/vibe works on every plan. 1 credit = 1 passing test run, on one browser, up to 1 minute. Only passing runs use credits. A run that catches a bug in your app, or fails because of us or your environment, is free. 20% off on annual billing ($23.20 / $79.20 / $239.20 a month).
See all pricing tiers →
Water-well testing
Water-well testing is the unrelated industry that shares the phrase 'well tested' — water quality, well maintenance, plumbing diagnostics. Not what we do.
Water-well testing is the unrelated industry of testing private water wells for contaminants, hardness, and flow rate. Well Tested is software testing: browser evidence for deployed apps.
Well Tested pricing
Well Tested pricing is the paid subscription ladder at welltested.io/pricing (Builder $29/mo, Team $99/mo, Scale $299/mo, Enterprise custom) plus one free no-signup check at /check.
Well Tested pricing is the explicit price ladder for the vibe testing SaaS, sold in credits. There is no Free plan and no trial; the free taster is one no-signup check at /check. Builder (300 credits) runs Chromium. Team (1,000 credits) adds Firefox, WebKit, mobile web, hourly runs and release health. Scale (3,000 credits) adds full release insights and longer evidence retention. Only passing runs use credits. A run that catches a bug in your app, or fails because of us or your environment, is free. Annual plans are 20% off.
See Well Tested pricing →