Click through a flow in your browser. AztecTests writes it out as a runnable Playwright or Robot Framework project — with self-healing locators baked in, so a UI tweak doesn't turn into a red pipeline.
No selectors to hand-write, no glue code to maintain.
The extension captures each step with ranked locators and an element fingerprint. Add checks inline as you go.
Choose Playwright (TS/JS/PY) or Robot Framework and download a project that runs after unzip — nothing to wire up.
When a locator drifts, the baked-in runtime recovers via ranked fallbacks and fingerprint matching.
Every element carries a stack of locators plus a fingerprint. Top one breaks, the next takes over — and it ships inside your exported scripts, not a service you rent.
Replay any recording in a real browser and step through screenshots with a pass / healed / fail badge on every step — before you ever export.
The same test becomes Playwright TypeScript, JavaScript, Python, or a Robot Framework suite. Switch stacks without re-recording.
A zero-dependency app that runs on your machine. Your app, your credentials, your recordings — none of it leaves.
The person who understands the application under test best isn't the AI — it's your QA. AztecTests is a QA assistant: it makes their work faster and more efficient, not a replacement.
Ask AI to write a test in plain language; it drives a real browser to find the elements and records a grounded test for you.
Point it at a folder and it proposes the edge, negative, and security tests you didn't write — then generates the ones you pick.
Turn the click-throughs you already do by hand into a self-healing test suite in the framework your team already runs.
Your QA understands the app best — AztecTests just helps them work faster and more efficiently, it doesn't replace them.