django-architect
Architect-level Django guidance — app boundaries and project layout, ORM and migration discipline, settings/environment split, auth and permissions, DRF API layering, admin as an internal surface, and the queries that make Django slow. Use when designing or reviewing a Django/DRF product.
- License 许可
- MIT
- 最近更新
- 2026-11-13
- Stars
- ★ 0
- Forks
- 0
01 前言 Front Matter
- name
- django-architect
- license
- MIT
- metadata
- { "crusor-path": "website-architect/backend-development-architect/python-architect/django-architect", "crusor-domain": "website-architect", "crusor-origin": "original", "crusor-sources": "docs.djangoproject.com; www.django-rest-framework.org", "crusor-triggers": "django, drf, django orm, django migration, django admin, django-rest-framework", "crusor-verified-on": "2026-10-30" }
- description
- Architect-level Django guidance — app boundaries and project layout, ORM and migration discipline, settings/environment split, auth and permissions, DRF API layering, admin as an internal surface, and the queries that make Django slow. Use when designing or reviewing a Django/DRF product.
- compatibility
- Django 4.2/5.x, Django REST Framework, PostgreSQL
03 SKILL.md 正文
Django Architect
Django is a product framework: it ships the ORM, auth, admin and migrations. Architecture is about respecting its grain — apps as boundaries, the ORM as the only data path — rather than fighting it.
When to use
- Product web app with users, auth, admin and a relational core.
- Content/catalog/CRM-style domains where CRUD is the bulk of the work.
- A team that benefits from convention over assembly.
When NOT to use
- A small service or single-purpose API (see
flask-architect,fastapi-architect). - Async-heavy or realtime-first workloads — Django works, but the async story is secondary.
- A polyglot microservice where the ORM's coupling is unwanted.
Decision 1 — Apps as bounded contexts
project/ settings, urls, asgi/wsgi
apps/<context>/ models, services, api, admin, migrations
services/ use-cases; transactions live here
selectors/ read queries (query objects)
- One app per business context, not per technical layer.
- Cross-app imports go through a service/selector, never into another app's models.
core/commonstays tiny; a big shared app is coupling with a nicer name.
Decision 2 — ORM discipline
- Views/serializers do not build complex queries; selectors do, and they are tested.
select_related/prefetch_relatedare part of the design, not an optimisation pass — N+1 is the default failure.- Write paths use
transaction.atomic()in the service, not the view. - No raw SQL without a documented reason and a parameterised query.
- Model
save()stays dumb; side effects belong to services/signals you can see.
Decision 3 — Migrations
- Every schema change is a migration, reviewed like code.
- Data migrations separate from schema migrations.
- Backwards-compatible sequence for zero-downtime: add → backfill → switch → remove (never rename in place on a live table).
makemigrations --checkin CI so drift cannot merge.
Decision 4 — Settings and environments
settings/base.py+dev.py+prod.py; environment-specific values from env vars (django-environ/pydantic-settings).DEBUG=Falseby default; a missing secret must fail at boot, not fall back.ALLOWED_HOSTS,CSRF_TRUSTED_ORIGINS, secure cookies and HSTS set explicitly in production.- Static/media served by the proxy/object store, never by Django in production.
Decision 5 — API layer (DRF)
- Serializers validate input and shape output; they are not business logic.
- ViewSets only for genuinely resourceful endpoints; explicit APIViews for actions with rules.
- Permissions are declared per view and tested; object-level checks in the service or permission class.
- Pagination, filtering and ordering are global defaults, not per-view reinvention.
- Throttling at the edge; heavy exports/imports go to background jobs.
Decision 6 — Admin and auth
- Admin is an internal operational surface: restrict by staff, audit actions, never expose it as the customer UI.
- Custom user model from day one (
AUTH_USER_MODEL) — retrofitting it is a migration project. - Session vs token auth chosen per surface; API tokens scoped and revocable.
Failure modes at scale
- Fat views + fat serializers; no service layer, rules duplicated.
- N+1 in list endpoints discovered only in production.
- Migrations locked in a deploy; long transactions blocking the table.
- Signals causing hidden writes that nobody can trace.
- Settings drift between environments; prod is the only place it breaks.
Verification receipts
manage.py check --deployclean;makemigrations --checkin CI.- Query-count test on the hottest list endpoint (
assertNumQueries). - Permission matrix tests per endpoint (owner / other user / staff / anon).
- Migration plan reviewed for a live table change (expand/contract).
- Load test on the two heaviest read and write paths.
04 同仓库更多 Skills
interaction-patterns-architect
Architect-level interaction design in code — choosing established patterns over invented ones, feedback and latency behaviour, destructive-action safeguards, navigation and overlay conventions, and the state model behind every interactive surface. Use when a screen has to behave, not just look.
design-system-ops-architect
Architect-level operation of a design system in code — token pipelines and versioning, component API and variant policy, the catalogue as the contract, adoption metrics, and the governance that stops a system from forking into three. Use when a component library must survive several teams.
design-handoff-architect
Architect-level design handoff — what must be specified before implementation starts, tokens over raw values, responsive and content-limit contracts, state matrices, and the review loop that catches drift between design and code. Use when a design is about to become code.
content-microcopy-architect
Architect-level UI content — where strings live, voice and terminology consistency, error and empty-state copy as part of the interface, translation-ready authoring, and the content model a product needs before it multiplies screens. Use when the words in the interface matter as much as the pixels.
ui-ux-implementation-architect
Architect-level direction for turning a design into working UI — the handoff contract, component ownership versus screen-level layout, interaction and state conventions, design QA gates, and the review habits that keep an implementation faithful without frozen pixels. Use before any specific UI implementation skill.
test-strategy-architect
Architect-level test strategy — allocating assertions across unit, integration, contract and end-to-end layers, designing for testability through dependency boundaries, fixture and data strategy, and the flake and coverage policies that keep a suite trustworthy. Use when defining how a codebase is tested.