The test
library.
Every entry below comes from a published Proofrun record. Open one to inspect the latest published prompt, frozen requirements, measurement boundary, version lineage, and every report that used it.
Sunwell Sanctuary
Large voxel-scene composition, landmark hierarchy, circulation, material separation, terrain integration, camera choice, and validator compliance.
Triage Desk
A multi-step support workflow: search and filters, synchronized state, validation, resolution, responsive operation, and accessible controls.
Weather Story
Accurate representation of supplied values, visual progression, honest demo-data labeling, responsive presentation, and meaningful interaction.
Verdant Terminal
A layered botanical station that tests architectural repetition, readable interiors, transparency, and camera discipline.
Orbit Radio
Interactive control design, coordinated UI state, responsive presentation, and the clarity of a compact one-screen instrument.
Prism Garden
Interactive creative coding, motion, variation, feedback, state counting, reset behavior, and visual polish in one offline artifact.
Sky Archive
Voxel composition, exact object-count constraints, visual hierarchy, connecting routes, material contrast, detail, depth, and camera framing.
Gravity Desk
Immediate playability, canvas physics behavior, active controls, object-count accuracy, responsive presentation, and visual craft.
A library is not a leaderboard.
This page describes the conditions Proofrun has published, not a claim that the catalog is complete or equally exercised. Quick web work still dominates the archive, Standard coverage is young, and no real Deep condition has been published.
Versions remain attached to their historical reports. A newer definition does not retroactively change an older run.