ce-dogfood-beta

Automated end-to-end browser QA runs by diffing the active branch against main to map changed user journeys, then driving browser-based test execution. It orchestrates diff analysis, flow mapping, and matrix construction to validate behavior autonomously.

Autonomous browser QA fixes failing tests by applying code fixes, adding regression tests, and committing each fix iteratively. This loop continues until the entire end-to-end test matrix is green and all edge paths pass.

Diff-scoped dogfood testing requires a running dev server, agent-browser, and access to branch diff outputs. These prerequisites enable the workflow to map user flows and execute the browser test matrix against the active branch.

Diff analysis maps changed user journeys by comparing branch differences against main to identify altered end states and flows. It builds an exhaustive browser-based test matrix from these findings, covering full journeys including edge and error paths.

The dogfood report documents diffs, mapped flows, applied fixes, and learnings generated during automated browser testing. It provides a comprehensive record of the end-to-end QA pass, detailing how changes altered user journeys and what was fixed.

Agent-browser drives tests against a running dev server to collect evidence and outcomes across full user journeys. It handles edge and error paths within the constructed test matrix to validate end-to-end behavior thoroughly.