dbt-tools-web-e2e-fix

To fix flaky Playwright E2E tests automatically, run an iterative fixer loop from the repository root using pnpm. This loop debugs failures with traces and screenshots, applies minimal targeted fixes, and re-runs tests until they pass, limited to five iterations.

To run Playwright E2E tests, you need pnpm and Node installed as specified in the repository, along with Playwright included via devDependencies. You must execute the test commands from the repository root to ensure the environment matches the project configuration.

Playwright E2E tests time out due to application delays or selector issues. You can debug timed-out tests by using traces, screenshots, and selective replays to identify the bottleneck, then applying minimal changes and re-running until the tests pass.

The best way to debug failing E2E tests while preserving test integrity is to use Playwright's built-in traces, screenshots, and selective replays. This approach guides minimal, targeted fixes rather than altering the core test logic to force a pass.

The main limitation of this automated E2E test fixing loop is the strict cap of five iterations. If the Playwright tests remain flaky, failing, or timed-out after five targeted fix and re-run cycles, the automated loop terminates without achieving a green state.