AI analysis
A SaaS screen for test running an AI automation workflow on sample data, shown in default, input picker, running, and step detail states.
Below a left sidebar navigation and a top breadcrumb sits a selector bar for workflow, version, input, and environment, and the body splits into a step-by-step run timeline, a right-hand assertions panel, and a run history table at the bottom. Blue primary buttons sit on a white background with light gray borders, passed, warning, and failed states are distinguished by green, orange, and red badges, and numbers and JSON are set in monospace. Each step row consists of an icon, a description, and a horizontal Gantt-style duration bar, while the input dropdown groups samples, live run replays, synthetic generation, and upload. The running state shows a streaming log panel, and clicking a step opens a right slide-over panel with latency, token, cost, and model metrics, output JSON, and a threshold warning. It is a useful reference for state design and information density in automation and AI pipeline debugging interfaces.
minimal · light-mode · corporate · techy
claude-opus-5-5 · 2026-11-04 09:16 · Confidence 0.92