What makes a platform agentic?

A platform is agentic when the planner, executor, and reviewer all live in one tool and the agent drives the workflow under explicit human approval. A patchwork of separate scripts is automation, not agency.

Is this just CI-triggered Playwright?

Yes for the execution; agentic for everything else. Well Tested grounds the plan in the live page, pins each approved test to an unchangeable version, heals steps inside the run with a label, triages failures with Jev + Clef, and returns replayable evidence.

Do I need to install anything?

No. Well Tested is a hosted platform. Browsers run in our infrastructure; you paste a URL and the planner runs. To test localhost, welltested tunnel --port exposes your dev server to our browsers, or welltested run --local runs kept tests in a browser on your own machine.

Where does the data live?

Runs and evidence live in your isolated workspace. Public share pages use revocable tokens. Private runs require explicit authentication and are never indexable.

Platform

An agentic testing platform you can actually trust.

One product: grounded planner, real-browser execution, durable evidence, revision-bound approvals. No parallel tooling to glue together.

What the platform includes

A grounded planner. Steps only target elements the scanner actually saw on the page.

Approved versions. Each approval locks one version of the plan, so reruns test exactly what you approved. Every approval, invalidation and repair decision is appended to an immutable audit log.

Durable evidence. Screenshots per step, failure video, browser console and network logs, and a diff against the last green run. Export any test as standalone Playwright code.

Autonomous browser testing

Approve a test once, then rerun it in real browsers on every GitHub push, every deploy (deploy hook URL) or a schedule, with result emails. A CI endpoint with exit codes and evidence links covers GitLab CI and Bitbucket Pipelines. Chromium runs on every plan; Firefox, WebKit and the mobile-web profiles run on Team and Scale.

Browsers included on each plan
Browser	Builder	Team	Scale
Chromium (desktop)	Included	Included	Included
Firefox and WebKit	Not included	Included	Included
iPhone 16 family and Pixel 9 mobile-web profiles	Not included	Included	Included

The phone rows are mobile-web profiles in a real browser engine. Native iOS and Android app testing is on the roadmap.

Your coding agent checks its work

npx welltested setup adds the Well Tested MCP server and a verify-before-done skill to Claude Code, Cursor, Codex, Windsurf, Copilot or Kiro. The agent reruns your kept tests (or drafts a plan you approve) before it says done, and gets the triage reason and a fix prompt back when one fails.

What the AI does

AI plannerLive
Plans from one sentence
Paste a URL and say what should work. The agent reads your live page and proposes steps that target elements it actually found. Nothing runs until you approve the plan.
Jev + ClefLive
Bug or not? AI triage
When a run fails, Jev reads the step trace and Clef checks the screenshots against your goal. You get a call: a real bug in your app, or a test or environment problem, with the signals it's based on. Your pass/fail result never changes.
Repair modelLive
Self-healing steps
When a button is renamed or moves, the run repairs that step on the spot from a fresh look at the page and keeps going. The step is labelled "Passed after repair" and you can keep the repair with one click. Assertions and destructive steps are never repaired.
Vision modelLive
Screenshot review
Screenshots from the run are reviewed against your goal and summarized next to the step trace, as an opinion separate from the pass/fail result.
PlaywrightLive
Real browsers
Chromium on every plan. Firefox, WebKit and iPhone and Android mobile-web profiles on Team and Scale. Native iOS and Android apps are on the roadmap, not live.
Speech to textLive
Voice to test
Hold the mic and describe the journey, up to 60 seconds in English or Turkish. The transcript becomes an editable goal; nothing runs until you approve it.

Frequently asked

What makes a platform agentic?: A platform is agentic when the planner, executor, and reviewer all live in one tool and the agent drives the workflow under explicit human approval. A patchwork of separate scripts is automation, not agency.
Is this just CI-triggered Playwright?: Yes for the execution; agentic for everything else. Well Tested grounds the plan in the live page, pins each approved test to an unchangeable version, heals steps inside the run with a label, triages failures with Jev + Clef, and returns replayable evidence.
Do I need to install anything?: No. Well Tested is a hosted platform. Browsers run in our infrastructure; you paste a URL and the planner runs. To test localhost, welltested tunnel --port <n> exposes your dev server to our browsers, or welltested run --local runs kept tests in a browser on your own machine.
Where does the data live?: Runs and evidence live in your isolated workspace. Public share pages use revocable tokens. Private runs require explicit authentication and are never indexable.

An agentic testing platform you can actually trust.

What the platform includes

Autonomous browser testing

Your coding agent checks its work

What the AI does

Plans from one sentence

Bug or not? AI triage

Self-healing steps

Screenshot review

Real browsers

Voice to test

Frequently asked

Related reading