Public specimen
Proof Studio / System surface lab
A stronger proof surface than one fixture .
The port comparison is useful, but it is small. This lab starts with the public specimen switcher and then records the deeper retained compositor evidence: eight UI-kit adapters, generated stories, and cross-kit comparison surfaces queued for publication.
Live public surface
Pick a view. The real specimen opens in the frame.
This is the public surface already on kbpd.ai: brand, motion, typography, density, and section choices resolve into the same component grammar.
Open the same navigation, button, card, and status grammar under the day register.
Open this exact viewInspect motion rails and timing decisions without changing the component contract.
Open this exact viewCompare value hierarchy and semantic color roles across the dawn register.
Open this exact viewReview how surface, border, elevation, and interaction tokens hold the page together.
Open this exact viewCompositor gallery
Openable public surfaces first; retained runtime traces beside them.
The stronger proof is not one berth screen. It is a set of surfaces that show what is public today, what is retained for audit, and what must be published before the multi-kit claim gets stronger.
public surface
Public component grammar
- Surface locus
- kbpd.ai specimen
- Observed proof
- Navigation, buttons, cards, status chips, and surface tokens resolve in one live specimen frame.
- Buyer use
- Open this first when judging whether the system has a coherent product grammar.
Evidence handles
- The surface is live on kbpd.ai, not a static screenshot.
- The same specimen route can switch sections and registers without leaving the page.
- The output is inspectable in the public browser frame.
Boundary: This proves the public specimen route survives register changes; it does not expose the retained multi-kit build.
- Surface locus
- kbpd.ai specimen
- Observed proof
- Motion rails, timing choices, and reduced-motion equivalents are visible in the same specimen system.
- Buyer use
- Use this row to judge whether motion is a product state language rather than decoration.
Evidence handles
- The public route opens directly to motion decisions.
- The register changes without replacing the specimen frame.
- The proof remains visual and interactive rather than hidden in JSON.
Boundary: This is a public motion specimen. It is not a measurement of user comprehension or conversion.
- Surface locus
- Proof Studio lab
- Observed proof
- Seven cold/backed public microsite pairs expose twenty-one measured visual, voice, product, and layout values.
- Buyer use
- Use this row when the question is whether the demo corpus is actually openable and reviewable.
Evidence handles
- Color, composition, voice, grid, perception, product, and typography pairs are public.
- Each pair has cold and backed outputs plus measured values.
- The gallery keeps limitations beside the values.
Boundary: The gallery proves artifact quality deltas on bounded fixtures; it does not claim market lift.
- Surface locus
- Proof Studio lab
- Observed proof
- Five cold runs sit beside five system-backed runs with mean-word, semantic, dependency, and self-sufficiency checks.
- Buyer use
- Use this row when the question is whether cold-start comparison has a visible matrix behind it.
Evidence handles
- The pilot matrix is public and navigable.
- The page keeps the preregistered benchmark boundary visible.
- The values are presented before the broad claim.
Boundary: This is calibration evidence. It stays below benchmark strength until the controlled benchmark clears.
Compositor runtime The retained integration check exercises status, style sources, catalogs, search/filter, screens, scenarios, design-system, and source-map paths.
Use this row to understand that the stronger compositor proof is a running system boundary, not only generated page screenshots.
- Surface locus
- Compositor runtime
- Observed proof
- The retained integration check exercises status, style sources, catalogs, search/filter, screens, scenarios, design-system, and source-map paths.
- Buyer use
- Use this row to understand that the stronger compositor proof is a running system boundary, not only generated page screenshots.
Evidence handles
- Status checks require loaded catalogs and style namespaces.
- Catalog checks cover component fields, button search, category filtering, and kit filtering.
- Screen generation checks happy, negative, and empty scenarios plus invalid-scenario validation.
Boundary: The trace is summarized publicly. The raw local runtime, environment, and repo paths remain retained until publication hygiene is complete.
Compositor plugin The retained build emits the plugin shell, UI frame, sandbox code, and Prime icon asset used by the Figma-style compositor.
Use this row to judge whether there is an actual built artifact behind the public proof roadmap.
- Surface locus
- Compositor plugin
- Observed proof
- The retained build emits the plugin shell, UI frame, sandbox code, and Prime icon asset used by the Figma-style compositor.
- Buyer use
- Use this row to judge whether there is an actual built artifact behind the public proof roadmap.
Evidence handles
- The retained dist output includes index, UI, and sandbox bundle artifacts.
- The changelog records a three-layer architecture: sandbox, UI, and middleware.
- The plugin UI work is source-driven rather than hardcoded visual styling.
Boundary: This public row is an artifact summary, not a downloadable plugin release.
Component-system adapter set Eight retained UI-kit adapters and 596 generated stories are recorded as the next publication target.
Use this row to see why the harbor fixture is not the strongest proof family.
- Surface locus
- Component-system adapter set
- Observed proof
- Eight retained UI-kit adapters and 596 generated stories are recorded as the next publication target.
- Buyer use
- Use this row to see why the harbor fixture is not the strongest proof family.
Evidence handles
- The adapter set spans PrimeReact, MUI, ShadCN, Radix, Ant Design, Chakra UI, Fluent UI v9, and Mantine.
- The retained story count separates per-kit stories from cross-kit comparison stories.
- The publication target is the sanitized multi-kit runtime plus screenshots for cross-kit comparison rows.
Boundary: The counts are public-safe. The generated story files are not yet published as a public runtime.
Retained compositor evidence
The larger proof program is the multi-kit compositor.
This is why the lab should not over-index on the port fixture. The stronger story is that structured system knowledge can keep surfaces coherent across brands, motions, component systems, and review modes.
Brand, motion, typography, and density can be changed from the specimen drawer.
Four deep links open live kbpd.ai surfaces with different section and register choices.
PrimeReact, MUI, ShadCN, Radix, Ant Design, Chakra UI, Fluent UI v9, and Mantine adapters are present in the retained compositor source.
The retained Storybook compositor output contains 545 per-kit stories and 51 cross-kit comparison stories.
Eight retained adapters
One specimen contract, many component ecosystems.
- 01 PrimeReact
- 02 MUI
- 03 ShadCN
- 04 Radix
- 05 Ant Design
- 06 Chakra UI
- 07 Fluent UI v9
- 08 Mantine
What this proves
System survival is visible; market lift is not claimed here.
Whether one retained system can keep a surface coherent while the visual register, motion register, section, and component evidence change.
A conversion claim, a universal accessibility pass, or a promise that every retained compositor artifact is already public.
Publish the sanitized multi-kit runtime, add screenshots for the cross-kit comparisons, and promote the best rows into the evidence desk as full case studies.