菜单
Provides expert-level Django development patterns covering App Router (indirectly via REST/GraphQL), async views, DRF, Celery, signals, middleware, and performance optimization. Use when building complex Django 5.x applications or identifying N+1 query issues.
默认使用会先检查来源的 Prompt；你也可以切换为直接命令，或下载本地副本。
决定是否安装前，请先阅读 SKILL.md，以及 SkillsMP 当前展示的配套文件。
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
直接命令不会经过审查 Prompt；运行前请先检查来源。
npx skills add https://github.com/tranhieutt/software_development_department --skill django-patterns命令会保持在同一行。复制前请横向滚动并检查完整内容。
想先保存到本地？可下载 SkillsMP 当前能够提供的文件。
Turns real agent failures, repeated prompts, team-specific workflows, and durable project lessons into better SDD skills or memory entries. Use when the user asks to create/update/refine skills, extract reusable lessons, improve skill routing, encode team process, or save patterns for future sessions.
Provides the vendored agent-style v0.3.5 prose rule pack as a portable Claude skill. Use when installing, syncing, applying, or auditing SDD Agent-Style support in another project, or when a style-review workflow needs the bundled rules.
Reviews Markdown prose against the pinned agent-style v0.3.5 rules and optionally writes a polished copy beside the source. Use for specs, ADRs, PR bodies, release notes, and technical docs when prose quality or evidence discipline matters.
基于 SOC 职业分类
正在显示 SKILL.md
| name | django-patterns |
| type | reference |
| description | Provides expert-level Django development patterns covering App Router (indirectly via REST/GraphQL), async views, DRF, Celery, signals, middleware, and performance optimization. Use when building complex Django 5.x applications or identifying N+1 query issues. |
| paths | ["**/*.py","**/manage.py","**/settings.py","**/urls.py","**/serializers.py"] |
| effort | 3 |
| allowed-tools | Read, Glob, Grep, Write, Edit, Bash |
| user-invocable | true |
| when_to_use | When building Django 5.x applications requiring async support, background tasks (Celery), real-time features (Channels), or advanced ORM optimization. |
select_related (FK) / prefetch_related (M2M) — never iterate and query inside loopsget_or_create race condition: wrap in transaction.atomic() in concurrent environmentssave() inside pre_save signal — causes infinite recursion; use update_fieldsbulk_create skips signals and save() — don't use when signal logic is requiredRunSQL with CONCURRENTLY index creation to avoid locks# FK / OneToOne → select_related (JOIN)
books = Book.objects.select_related("author", "author__publisher").all()
# M2M / reverse FK → prefetch_related (separate query)
authors = Author.objects.prefetch_related("books", "books__tags").all()
# Custom prefetch with queryset
from django.db.models import Prefetch
Author.objects.prefetch_related(
Prefetch("books", queryset=Book.objects.filter(published=True), to_attr="active_books")
)
from django.db.models import Count, Avg, Q, F, Value
django.db.models.functions Coalesce
Author.objects.annotate(
book_count=Count(),
avg_rating=Coalesce(Avg(), Value()),
high_rated=Count(, =Q(books__rating__gte=)),
).(book_count__gt=).order_by()
# BAD — race condition
product = Product.objects.get(pk=pk)
product.stock -= quantity
product.save()
# GOOD — atomic at DB level
Product.objects.filter(pk=pk).update(stock=F("stock") - quantity)
from django.views import View
from django.contrib.auth.mixins import LoginRequiredMixin
class OrderDetailView(LoginRequiredMixin, View):
def get(self, request, pk):
order = get_object_or_404(Order.objects.select_related("user"), pk=pk, user=request.user)
return JsonResponse(OrderSerializer(order).data)
class ProductSerializer(serializers.ModelSerializer):
class Meta:
model = Product
fields = ["id", "name", "price", "stock"]
read_only_fields = ["id"]
def validate_price(self, value):
if value <= 0:
raise serializers.ValidationError("Price must be positive.")
return value
def validate(self, data): # cross-field
if data["stock"] == 0 and data.get("is_featured"):
raise serializers.ValidationError("Out-of-stock products cannot be featured.")
return data
from rest_framework import viewsets, status
from rest_framework.decorators import action
from rest_framework.response import Response
class ProductViewSet(viewsets.ModelViewSet):
queryset = Product.objects.select_related("category")
serializer_class = ProductSerializer
permission_classes = [IsAuthenticated]
@action(detail=True, methods=["post"], url_path="archive")
def archive(self, request, pk=None):
product = self.get_object()
product.is_archived = True
product.save(update_fields=["is_archived"])
return Response(status=status.HTTP_204_NO_CONTENT)
# settings.py
REST_FRAMEWORK = {
"DEFAULT_FILTER_BACKENDS": ["django_filters.rest_framework.DjangoFilterBackend"],
"DEFAULT_PAGINATION_CLASS": "rest_framework.pagination.CursorPagination",
"PAGE_SIZE": 20,
}
# viewset
class ProductViewSet(viewsets.ReadOnlyModelViewSet):
filterset_fields = {"price": ["gte", "lte"], "category": ["exact"]}
ordering_fields = ["price", "created_at"]
search_fields = ["name", "description"]
from celery import shared_task
from django.db import transaction
@shared_task(bind=True, max_retries=3, default_retry_delay=60)
def send_order_email(self, order_id: int):
try:
order = Order.objects.select_related("user").get(pk=order_id)
send_email(order.user.email, order)
except Order.DoesNotExist:
return # don't retry if deleted
except Exception as exc:
raise self.retry(exc=exc)
# Dispatch after DB commit — avoids race condition
def create_order(data):
with transaction.atomic():
order = Order.objects.create(**data)
transaction.on_commit(lambda: send_order_email.delay(order.pk))
| Pitfall | Fix |
|---|---|
| N+1 queries in serializer | select_related / prefetch_related on queryset |
objects.all() in views | Always filter + limit; never expose unbounded querysets |
| Storing secrets in settings.py | Use django-environ or environment variables |
DateTimeField(auto_now_add=True) not testable | Use default=timezone.now for overridable defaults |
| Sync ORM in async views | Use sync_to_async or Django 4.1+ async ORM |