Skip to content

v0.10.0 — Domain stories & design system driving UI mocks

Choose a tag to compare

@wfukatsu wfukatsu released this 24 Jun 08:24
· 334 commits to main since this release
e1f2b7f

All three plugins (product, architect, scalardb) are released together at 0.10.0.

Highlights

This release makes the product UX pipeline produce concrete, consistent screens: domain stories decide what each screen does and in what order, and a separately-managed design system decides the shared visual language the mocks render in.

persona → journey → positioning → create-domain-story → design-system → generate-ui-mock → ...
                                   (what / order)         (how it looks)   (story + tokens → screens)

Added

  • /product:create-domain-story — persona-anchored Domain Storytelling. Actors come from personas (PER-), activities from job stories (JOB-) ordered by the journey (JNY-), work items from the things handled. Each story is the chosen happy-path scenario for a persona pursuing a key job (scoped per persona×job; bounded contexts are optional enrichment via --domain). Runs after journey/positioning and before UI mocks; outputs reports/01_ux/domain-stories/ with STORY- traceability. The product-pipeline counterpart of /architect:create-domain-story.
  • /product:design-system — build or --import a separately-managed design system. Build derives W3C DTCG tokens (color/type/spacing/radius/elevation/motion) from positioning/personas/vision with a WCAG contrast gate; --import normalizes an existing system (Tailwind config / DTCG JSON / Figma Tokens / CSS theme) into the same schema. Output lives in a dedicated design-system/<name>/ tree (not under reports/), carries a semver manifest.json, supports multiple coexisting named systems, and is standalone. The active system is recorded in work/pipeline-progress.jsonoptions.design_system. New rule rules/product/design-system.md. The product plugin now has 24 skills.

Changed

  • /product:generate-ui-mock is story-driven and design-system–styled — screens are derived from the domain story for each persona×job (one activity ≈ one screen interaction), and styled by the active design system: its tokens.css is injected into every self-contained screen, rendered at --fidelity=lo (tokens only) or mid (tokens + CMP- component styles). Falls back to ad-hoc lo-fi styling when no system is configured. Screens now also trace STORY-/CMP-.
  • UX-phase orderingcreate-domain-story and design-system run after positioning and before generate-ui-mock in the full profile.

Full changelog: CHANGELOG.md · 日本語: CHANGELOG_ja.md