// no-code test automation
No code. No recording. Just a URL.
Other "no-code" testing tools still need you to record yourself clicking. AegisRunner crawls the whole site, writes the Playwright tests, and runs them on every deploy — without any input from you beyond the URL.
// what you do
Four steps — only one needs your hands.
Paste a URL
That's the input. No SDK, no recording, no test plan.
AI explores the site
Real browser, real interactions — forms, logins, modals, pagination. Roughly 1 minute per 20 pages.
You get the report
Plain-English list of what works and what's broken. The Playwright test code is generated alongside.
Schedule it
Daily / weekly / per-deploy. AegisRunner re-runs and pages you if anything regresses.
// vs record-and-playback
"No-code" means a lot of different things. Here's what ours means.
| Dimension | AegisRunner | Record-and-playback tools |
|---|---|---|
| Where does the test come from? | AI writes it from your URL | You record yourself clicking |
| Need to set up tests per page? | No — crawls the whole site | Yes — one recording per flow |
| What if the UI changes? | Auto-heal repairs the selector | Recording breaks — re-record |
| Cross-browser? | Chromium / Firefox / WebKit | Often Chrome-only |
| Code export? | Plain Playwright TypeScript | Proprietary format |
| Hidden pages discovered? | Yes — the AI crawls every link | No — you only test what you record |
// faq
Frequently asked
How is this different from record-and-playback tools (Mabl, Testim, Reflect)?
Record-and-playback still requires a human to click through every flow you want to test. AegisRunner needs only a URL — the AI does the exploration and writes the tests. Coverage scales with site size, not with QA hours.
I don't code at all. Can I really use this?
Yes. The whole flow is browser-based, no installation. You see results as a plain English report ("Sign-up works", "Cart total is wrong on mobile") with screenshots — the Playwright code underneath is optional and only matters if you ever want to take the tests with you.
Can I customise what gets tested?
You can add custom prompts ("Always test the cart icon shows the count", "Never click the Delete button"), include / exclude URL patterns, configure login scripts. Defaults are sensible — most teams never touch these.
What if the AI generates a wrong / generic test?
Edit it. Every step has a UI editor — change actions, selectors, expected values. Or click "Regenerate" with a custom prompt. Or just delete it. The AI is the first draft; you stay in control.
Try it on your URL — free
No card, no install, no recording. First scan in about a minute.