Your tests, your CI
Tests live in your repository and run where they already run. A zero-dependency client sends every run's results — nothing to host, nothing to migrate.
OpenAssert collects the results of your own end-to-end tests and judges each run against the Then assertions you declared. Failures become tickets, and your coding agents can fetch them and link the fix themselves.
What it is
OpenAssert does not run your tests. Your project keeps running its own Playwright suite in CI; OpenAssert keeps the record of what was asserted, what held, and what still needs fixing.
Tests live in your repository and run where they already run. A zero-dependency client sends every run's results — nothing to host, nothing to migrate.
A run passes only when every declared Then assertion passed and the tested versions were recorded. The dashboard shows, per environment, which assertions hold.
Every failing case is tracked as a bug with three states: needs a ticket, in progress, or fixed — so nothing that broke is quietly forgotten.
People and their coding agents sign in from the terminal, fetch the current failures, and link tickets themselves. The loop closes without copy-pasting reports.
How it works
The same Given / When / Then cases your team already reviews become the contract every release is checked against.
Write each case as Given / When / Then and implement it as a Playwright test in your repository.
Run the suite in CI as usual. The client reports every run, with each assertion's result and the versions under test.
See which assertions hold in each environment. Failing cases become tickets that you — or your coding agent — pick up and link.
It grows out of an assertion-checking tool already used to ship real projects. Early access is not open yet — follow the repository to hear when it is.