Orbis QAOrbis QA

Describe a test.
Watch it run.

Orbis QA reads a scenario in plain English and runs it in a real browser. No selectors. No fixtures. No cloud. Bring your own LLM key, point it at your app, and walk away.

Download for freeRead the docs
Platforms
3
macOS · Win · Linux
License
Apache
2.0, open source
Account
None
No sign-up
scenarios/checkout.txtScenario
Go to https://shop.example.com.
Add the first product to the cart.
Open the cart and verify the
total is greater than zero.
Check out as a guest with the
test card 4242 4242 4242 4242.
Confirm the order success page
shows an order id.
7 linesPlain text. No code.
What it does

Everything an SQA needs. Nothing extra.

See full feature list

Plain English scenarios

Describe the user journey in two sentences. The planner figures out clicks, types, and assertions step by step.

Reasoned recovery

When a step fails, a dedicated diagnosis call decides retry, skip, or abort. Not a brittle three-retry loop.

Suites

Group scenarios into named batches. Run the nightly smoke test with one command and aggregate the result.

Run vs. run compare

Side by side diff with drift highlighting. Spot what actually changed between yesterday and today.

Live progress and cost cap

Every step streams live with a screenshot. A per-run cents cap stops a runaway scenario from draining your key.

Flake detection

Pass rate per scenario over the last 20 runs. Real flake telemetry without setting up a dashboard.

How it works

Four steps. No selectors.

You describe what to test. The agent observes, reasons, acts, and asserts in a loop until your scenario is done, or it tells you why it could not be.

  1. 01
    Perceive

    Snapshot the page: ARIA tree, screenshot, recent network slice.

  2. 02
    Reason

    The planner picks the next action and assertion, with the scenario prose in context.

  3. 03
    Act

    Click, type, scroll, navigate. Driven through a real browser, not a headless simulation.

  4. 04
    Assert and recover

    Check assertions. If something fails, the recoverer diagnoses and decides what to do next.

The boring fundamentals

Built to live on your laptop, not in someone else's cloud.

On your machine

No cloud, no telemetry by default. Scenarios, runs, and screenshots all live in your data folder.

Bring your own LLM

Anthropic, OpenAI, Gemini, DeepSeek, xAI, Mistral, OpenRouter, Ollama. The key sits in your OS keychain.

Hardened by default

Sandboxed screenshot protocol, locked DevTools in production, error scrubbing, integrity-checked bundle.

Start running scenarios in under a minute.

Download Orbis QA, paste your LLM key, describe your first test. No account. No tracking.

DownloadGetting started