Private frontend workshop, shown through the artifact

A living proof surface for sites, agents, and product labs.

JD Workshop is Joshua David Sta Rita’s build room: fast frontends, motion with restraint, labeled concept work, and working demos that make the claim visible before the copy explains it.

  • 01 isolated site
  • 02 guide route
  • 03 critique loop
01 / Proof floor

Receipts before rhetoric.

The current JD Workshop repository stays the factual base. This pilot expands the surface, but every card tells you what kind of proof you are looking at.

Real

Current JD Workshop repo

Next.js app, profile route, motion layer, schema split, selected-work cards, and a tokenized visual system from the active repository.

Visit live site
Real

Frontend proof language

Strict TypeScript, React/Next.js, Vue/Nuxt, performance discipline, reduced motion, and public GitHub presence form the technical baseline.

Open GitHub
Concept

Operator intake cockpit

A fictional service-business cockpit with synthetic lead data. It shows the kind of AI workflow surface JD Workshop can build, not a shipped client result.

Built demo

Proof labeling system

This pilot implements the evidence labels in the interface itself. The guide route explains what is real, concept, and demo material.

View guide
02 / Concept lab

Fiction can be useful when the label is honest.

Some variations in the full 25-site run should invent realistic apps, brands, dashboards, and workflows. The trust rule is simple: concepts are named as concepts, then the strongest ones get built until they become real demos.

Concept

Field Notes CRM

Sample client: fictional. Data: synthetic. Purpose: demonstrate service-business workflow design.

Built demo target

AI quote triage agent

Can become a working form-to-summary workflow during the full run if the concept proves useful.

Real foundation

JD Workshop identity

Colors, type, links, slogans, and repo facts come from the active project files.

03 / Agent system

A website that can explain the workbench behind it.

The full run should not only display final pages. It should show how agents, automation, media generation, browser critique, and proof gates combine into a repeatable workshop system.

  1. BriefRepo foundation, owner steering, proof rules.
  2. GenerateStill and video atmospheres, optimized for web.
  3. BuildIsolated prototype with guide route and labels.
  4. CritiqueDesktop, mobile, console, responsiveness, fixes.
  5. ConvertPromising concepts become working demos.

04 / Decision surface

This pilot tests the rule for all 25 variations.

If it works, the full run can safely go wider: client-facing, hiring-facing, cinematic, strange, product-heavy, agent-heavy, and spectacle-driven, while keeping the evidence line visible.