How It Works
You have a REST API. You want to know it works — not just that individual functions return the right thing, but that the whole system behaves correctly when real services are running. That's what Spark does.
Spark is a test runner built for integration testing. You describe your test in a simple YAML file — which services to start, what request to send, and what the response should look like. Spark spins up real Docker containers, waits until they're healthy, fires your request, and checks the result. When it's done, everything gets cleaned up automatically.
No test frameworks. No mocking libraries. No glue code. Just YAML and Docker.
What Spark handles for you
Service orchestration
Spark starts Docker containers in isolated networks. Each test gets its own environment — no shared state, no interference between tests.
Health checking
Before running anything, Spark waits for every service to report healthy. No more race conditions or "connection refused" failures.
Execution
Send HTTP requests or run CLI commands against your services. Spark captures the full response — status codes, headers, body, stdout, stderr.
Assertions and reporting
Check that the response matches what you expect. Get HTML dashboards, JUnit XML for CI, or plain terminal output.