render-debug-loop

Automated visual regression testing uses a build-run-screenshot loop to capture and compare render outputs against reference baselines. This unattended workflow detects visual drift across zooms, camera offsets, and render modes to validate iterative rendering improvements.

Automated render-debug workflow repeatedly builds, runs, and captures screenshots to diagnose visual regressions. It evaluates outputs against topic-indexed baselines for trixels, SDF shapes, lighting, and backend parity to guide fixes until evaluation passes.

Configure a demo project with an auto-screenshot flag and a shot table to set up automated screenshot comparisons. The workflow builds via fleet-build, runs via fleet-run, captures screenshots across the shot list, and compares them against reference baselines.

Yes, automated render loops require a demo project configured with the auto-screenshot flag and a defined shot table. The workflow uses fleet-build and fleet-run commands to compile and execute the project for unattended visual validation.

Yes, visual drift detection captures screenshots across a demo's shot list covering different lighting, trixel shapes, and backend rendering modes. The system compares these captures against topic-indexed reference baselines to identify regressions.

Iterative rendering improvements use an automated loop that applies fixes, rebuilds, runs, and captures screenshots until evaluation passes. This reference-driven diagnosis guides regression remediation across trixels, lighting, and backend parity.