One cockpit
Plan, automate and run UI and API tests from a single workspace, with every case, run and change visible on one dashboard.
Coming soon
A local-first AI test cockpit for planning, generating and running UI and API tests.

One cockpit for planning, automating and running UI and API tests. The capabilities below are available in preview; the next section describes planned additions.
Available in preview
Plan, automate and run UI and API tests from a single workspace, with every case, run and change visible on one dashboard.
Describe a flow and point to a URL. QA OS walks the real page, including frames, shadow DOM and pop-ups, and writes Gherkin cases only from controls it actually found. Existing cases are reused as preconditions instead of being rewritten.
API requests recorded during a UI walk become API test cases in one click.
Generate positive, negative, boundary and injection cases from OpenAPI, Postman, cURL, PDF or Excel. Each case comes with a ready cURL and the schema it asserts.
Selected cases become a Playwright TypeScript suite. The suite is run once to prove it really executes, and failures are reported as the product's answer, not hidden as flaky scripts.
Run any folder against any environment, with consent for write requests. Every verdict carries its reason, and you can rerun only the failures.
When a control changes, QA OS proposes the exact fix side by side. Nothing is written until you approve it, and approved fixes are verified in a separate run.
Create cases by hand, reuse existing cases as preconditions, and validate structure and references before saving.
Every cockpit action is also a CLI command, and both give the same answer.
Runs on local models through Ollama. Your code and data stay on your machine.
Record a real browser session, then turn it into UI cases, API cases or a k6 / Gatling load test.
Versioned requirements and specs you can ask questions of, with cited answers and test plans tied to the exact source version.
When a pull request lands, QA OS maps the change to the tests that matter, runs them in the PR environment and reports on the merge check.
Connect GitHub, API specs and MCP tools such as Jira and Playwright once. Every new change then uses the same saved policy.