Planning corpus: this is the real design-discussion doc for epic demo-store-plant-shop, synced verbatim from this repo’s own internal planning process.

Design Discussion: demo-store-plant-shop

0. Context

Backlog epic 37. User’s explicit ask: a third demo store, “somewhat eclectic but a classic commerce store… an etsy neat plant store or something” (illustrative example, not literal — resolved below as a general eclectic handmade-goods shop with plants as one category among several, matching real small-batch Etsy-style sellers who rarely sell just one thing). Depends on epic 34 (done — switchable themes). Architecturally simpler than epic 36: this is a genuinely new demo, not a rename, and needs no bespoke new capability (no personalization field, no service-area model) — it exercises the exact registry extension point epic 36 just proved works cleanly (DEMO_SLUGS/DEMO_REGISTRY/defaultThemeKey, confirmed by reading apps/reference-storefront/lib/demos.ts directly before writing this doc).

1. Design questions

(a) Brand identity and category shape. Resolved: “Broadleaf & Co.”, slug broadleaf, an eclectic small-batch goods shop across 4 real categories — Plants, Ceramics & Planters, Textiles & Fiber Arts, Paper & Ephemera. This mirrors how real Etsy-style sellers actually operate (rarely one narrow category) and gives the “eclectic…classic commerce” framing real substance, not just a plant store with a different name.

(b) New seed module, or extend an existing one? Resolved: a new lib/seed-broadleaf.ts, mirroring lib/seed-northline.ts’s module shape (exported seedBroadleafDemo(deps) function, registered in DEMO_REGISTRY) — not a branch inside lib/seed.ts (print-shop’s module), keeping each demo’s content genuinely independent per this repo’s established one-module-per-demo convention.

(c) Does the nav/depth work need new code, or does it “just work”? Resolved, confirmed by direct inspection of app/demo/[demoSlug]/layout.tsx’s buildNavLinks (built by epic 35, proven again by epic 36): nav links are already built server-side per demo from real MarketingCatalogService category data — a demo with real categories automatically gets a real nav with zero new nav code. This epic’s “true menu, true set of products” critique is satisfied entirely by real seed-data depth, not new architecture. This materially shrinks this epic’s scope versus 35/36.

(d) Default theme. vibrant (existing bundle: warm orange/pink palette, playful 12px radius) — the one bundle not yet used as any demo’s default (northline→northline, print-shop→editorial), and a genuine fit for a colorful, eclectic craft-market identity.

2. Scope assessment

Medium. One new seed module, one registry entry, no new subsystem, no schema change, no new shared component work (nav/theme mechanisms already handle a 3rd demo for free).

3. Stories

new-demo-seed-and-registry — new lib/seed-broadleaf.ts with 4 real categories and >=8 real products (real names/descriptions/prices, no lorem ipsum), including at least one tiered-variant product (e.g. a plant sold in small/medium/large pot sizes, mirroring the generateSkus tiering pattern epics 35/36 already used), registered in DEMO_SLUGS/ DEMO_REGISTRY with defaultThemeKey: "vibrant".
verification-and-closeout — live-verify the new demo end to end (nav auto-populated with the 4 real categories, all product/PDP pages render, cart/checkout reaches the same expected Stripe-key-absent failure point as every other demo, fresh no-cookie visit shows the vibrant theme by default), confirm print-shop and northline are both completely unaffected, update README.md/VISION.md’s demo roster mentions, close out the backlog row, merge to master.

4. Risks

Low. This is the third instance of an already-twice-proven extension point (DEMO_SLUGS/DEMO_REGISTRY/defaultThemeKey) — epics 35 and 36 already demonstrated the nav/theme mechanisms generalize correctly to additional demos with zero shared-component changes.

5. Open questions

None blocking.