QA Arsenal

Autoquiver

Jalaran Autoquiver generates Playwright browser tests from a plain-language description and runs them on a schedule, tracking results across runs.

Open Autoquiver →Create free account

Who Autoquiver is for

For QA engineers who want end-to-end coverage of a critical path without hand-writing the selectors, and who want the suite to run overnight rather than when someone remembers. Read the limits first — execution is deliberately restricted.

What Autoquiver does

Describe a flow in plain language and Autoquiver generates the Playwright Python to exercise it. Suites can be scheduled to run on a recurring basis, with screenshots captured and pass or fail history kept across runs, and a Discord alert on failure so a broken build finds you rather than the other way around. Generated scripts pass an abstract-syntax-tree guard before they are stored and again before they run — no imports, no dynamic evaluation, nothing but the browser primitives a test actually needs.

Plain-language to Playwright Python generation
Scheduled recurring suite runs with failure alerts
Screenshots and pass/fail history across runs
Every script passes an AST allowlist before storage and before execution

How Autoquiver works

Describe the flow
Plain language. Autoquiver generates the Playwright Python that exercises it.
Review the generated script
It passes a code guard on save — an allowlist that rejects imports, dunder access and dynamic evaluation.
Schedule the suite
Recurring runs alongside the platform’s other scheduled work, with a Discord alert when a suite fails.
Read the run history
Pass and fail across runs with screenshots, so a flaky test is visible as flakiness rather than as one bad morning.

What Autoquiver does not do

Execution is owner-only, and this is the most important thing on this page. Autoquiver runs user-authored Python, and on a shared platform that capability cannot be handed to arbitrary accounts — so non-owner accounts can write and generate scripts but the run endpoints refuse them, and the interface says so rather than failing mysteriously. The code guard is a deliberate allowlist, not a full sandbox; nobody can build one of those. Concurrency is bounded to a small number of simultaneous suites.

Common questions

Can I run suites on my own account?

Not currently. Execution runs user-authored Python and is restricted to the platform owner for that reason. Non-owner accounts can create and generate scripts, and the interface tells you plainly rather than letting a run fail without explanation.

Is generated code executed without review?

It passes an AST allowlist when saved and again immediately before execution — no imports, no dunder attribute access, no dynamic evaluation. That is defence in depth on top of the owner restriction, not a substitute for it.

What happens when a scheduled suite fails?

A Discord alert fires if a webhook is configured, and the failure is recorded in the run history with screenshots so you can see what the browser saw.