Status: RATIFIED as working canon 2026-09-16 (implementation of lead G1 on docs/positioning-g1-lead; tracker #614). Prior leads retired in POSITIONING. Tag reviewed: v3.33.0. Tracker: icvoss/django-brickwork#614. Method: buyer job and package capability first; slogan last. Existing POSITIONING / site copy were cited only as drift evidence.

This document deliberately ignores "Beautiful interfaces for anything" and "Building blocks for beautiful apps and websites" as candidates until the end, where they are scored against the greenfield frame (not the other way round).

1. Buyer and job (ordinary workflow)

Actor: a Django team shipping real product surfaces on server-rendered templates (often HTMX + Alpine), not a React SPA team and not an admin-only team.

Trigger: they need a public site, an app shell, an ops queue, docs, auth, or content surfaces to look finished and share one brand, without rebuilding design decisions per project.

Job: install one interface foundation, brand it, compose or copy pages, wire Django behaviour, ship. Later: upgrade the foundation without rewriting owned pages.

Failure modes they already know:

Escape	Why it fails the job
Hand-roll Tailwind forever	No shared contract; every page reinvented
Admin skin (e.g. Unfold-class)	Wrong surface class for hand-built product UI
HTML/React Tailwind kit	Not Django-aware; no pin/upgrade story for shells
Copy-paste component dump	Looks like progress; breaks on redesign and a11y

Success: one branded system across the surfaces they ship this year; pages they own; foundation they can pin.

2. What the package actually is (capability, not aspiration)

Consumes

Django project, consumer routes/views/data, consumer brand CSS on --bw-*, documented Alpine/HTMX/Tailwind stack.

Does (three jobs in one wheel; ADR-108)

Job	Meaning at 3.33
Substrate	Tokens, 6 shells, 70 components, interaction contracts, a11y CI, Theme axes
Reference catalogue	40 archetypes + 32 sections + skeleton; copy and own (ADR-056)
Starter	`manage.py startsite` emits an owned project (ADR-095)

Plus a named Theme product (THEME.md): L1 Recolour through L4 Rhythm on the same tokens.

Produces

An installable, semver-governed UI foundation. Success is not "the consumer page is beautiful forever"; success is "the reusable interface decisions are owned here, brandable, tested, and upgradable."

Shipping coverage (catalogue-manifest at 3.33)

Interface family	Archetypes shipped	Notes
Product applications	13	Strongest app coverage
Documentation	7	Complete vs INTERFACE-SYSTEM required set
Editorial and publishing	7	Complete vs required set
Data-heavy operations	6	Material ops set; viz still contract-bound
Marketing and public web	4	Partial vs required marketing list
Transactional journeys	3	Partial (auth-heavy; checkout-class gaps remain)

Honest completion line: brickwork is a shipping interface foundation with deep app/docs/editorial coverage and partial marketing/transactional coverage, plus a Theme profile ladder. It is not "every interface archetype in INTERFACE-SYSTEM is done."

Stale: INTERFACE-SYSTEM "Current state" still says documentation/editorial are the highest-priority missing families. That paragraph is false at 3.33 and must be rewritten in the same wave as POSITIONING.

Craft evidence (not lead substitutes)

Meet-bar PASS (cited audit).
Beat scorecard: lead or tie on 6 of 8 surfaces; S5 Lose; S3 N/S. Soft-pass forbidden.
A11y: mechanism at 236 documents, not a guarantee.
No house aesthetic: brand packs are consumer-owned.

3. Category (greenfield)

Category name to own:

Server-rendered Django interface foundation (design system + shells + copyable pages + Theme profiles).

Not the category:

Django-admin theme
CMS / page builder
General Tailwind utility kit
"Tailwind for Django" (rejected as analogy-led identity)

Competitive frame (bounded): no package found combining app shell, components, theming and marketing kit for server-rendered Django (search-scope hedge required). Visual bar remains Tailwind UI kits for craft only.

4. Point of difference (what only this offer stacks)

Stack these; do not lead with inventory counts alone:

Django-native substrate that pins and upgrades under semver.
Pages you own (copy-paste examples off the loader path).
One brand across families via --bw-* and Theme L1 to L4.
Accessibility as CI mechanism at catalogue scale.
Craft aimed at kit bar with published miss honesty (S5).

The upgrade boundary (1+2) is the mechanism that makes giving the catalogue away rational. Theme (3) is the 3.33 product surface most undersold publicly. Coverage (families table) is proof of range, not a promise of "anything."

5. Positioning statement (classic form)

For Django teams building hand-built product and public interfaces who need one branded, accessible design system they can pin and upgrade, brickwork is the server-rendered Django interface foundation that ships shells, components, Theme profiles and copyable pages they own, unlike admin skins, HTML-only Tailwind kits, or endless hand-rolled UI, brickwork keeps the reusable interface contract in one upgradable package while consumer pages stay theirs.

6. Lead candidates (derived now; prior slogans scored later)

ID	Lead	Why it fits capability	Risk
G1	One interface system for the Django surfaces you ship.	Matches six-family coverage + Django-native category; "surfaces you ship" hedges "anything"	Needs family honesty nearby
G2	Pin the foundation. Own the pages.	Names the ADR-056/095 mechanism buyers actually buy	Underplays craft and Theme unless support lines carry them
G3	Theme your Django UI without forking the kit.	Centres 3.33 Theme product	Narrow if Theme is support, not lead

Recommended lead: G1.

Recommended support line (subordinate, not competing): G2's mechanism ("Pin the foundation. Own the pages.") plus a Theme proof line ("L1 recolour to L4 rhythm on --bw-*").

Recommended anti-leads: beauty as the organising noun; "anything" without a shipping matrix; inventory counts as the hero; competitor names as identity.

7. Score prior slogans against this frame (not as options to keep by default)

Prior line	Score vs greenfield	Verdict
Beautiful interfaces for anything.	"Beautiful" is aim+audit, not a certified consumer outcome; "anything" overclaims vs partial marketing/transactional	Retire as lead. May survive only as aspirational north-star prose inside INTERFACE-SYSTEM, never as public H1
Building blocks for beautiful apps and websites.	"Building blocks" undersells system/Theme/shells; sounds like a dump; "beautiful" still soft	Retire as lead. "Building blocks" may appear in subordinate mechanism copy if useful

Neither prior line is the greenfield recommendation. Choosing A vs B from

614's first draft is the wrong framing; choose G1 (or G2/G3 with

recorded reason) and rewrite POSITIONING from that.

8. Message hierarchy (what every surface must carry)

Lead (G1): one interface system for the Django surfaces you ship.
Category: server-rendered Django interface foundation (not admin, not CMS).
Mechanism: pin/upgrade substrate; own copied pages; Theme L1 to L4.
Proof: family shipping matrix; meet/beat cites with misses; a11y mechanism.
Boundary: consumers own data, permissions, domain; no house aesthetic.

9. What to rewrite when the ruling lands

Artefact	Change
`docs/POSITIONING.md` §1 to §3	Greenfield statement, G1 lead, shipping matrix instead of blanket "not yet complete"
`README.md` opening	Same lead; Theme named; family honesty
`docs/INTERFACE-SYSTEM.md` Current state	Docs/Editorial complete; remaining gaps named
brickworkui.com	Align H1/meta to ruling; remaster #233 proceeds on display
Umbrella marketing plans	Banner prior leads; point here

Claim-honesty rules (§9 style) and no-house-aesthetic (§10 style) carry forward as process rules even when slogans die.

10. Decision asked of the owner

Accept G1 as the public lead, or name G2/G3 / a rewritten G-line with equal capability grounding.
Explicitly retire both prior public leads (do not leave dual canon).
Authorise a POSITIONING + README + INTERFACE-SYSTEM current-state PR on that ruling before large site remaster copy locks.

Site display remaster may continue on wireframes that are lead-agnostic (show Theme, families, craft), but no H1/meta lock until this decision is written into POSITIONING.