ericrisco/rsc-harness / skills/django
django
VerifiedUse when building, reviewing, securing, testing or shipping a Django app — models, migrations, QuerySets/managers, FBV/CBV views, forms, the admin, settings split, and Django REST Framework (serializers, ModelViewSet, permissions). NOT async FastAPI/Pydantic services (that is `fastapi`), NOT Postgres schema/index work (that is `postgresdb`).
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 12.1 KB |
| evals/README.md | markdown | 901 B |
| evals/cases.yaml | config | 3.1 KB |
| references/drf.md | markdown | 4.0 KB |
| references/orm-performance.md | markdown | 3.1 KB |
| references/security.md | markdown | 3.0 KB |
| references/testing.md | markdown | 2.7 KB |
| scripts/verify.shexecutable | script | 4.3 KB |
references/orm-performance.md
View on GitHub# ORM performance recipes
The ORM is fast when you tell it what to fetch. Almost every slow Django page is an N+1 or a
fetch of columns/rows you never use.
## select_related vs prefetch_related vs Prefetch
```python
# Forward FK / OneToOne -> SQL JOIN, one query
Article.objects.select_related("author", "author__profile")
# Reverse FK / ManyToMany -> a second query, joined in Python
Author.objects.prefetch_related("articles", "tags")
# Prefetch that needs its own filter/order/select_related
from django.db.models import Prefetch
Author.objects.prefetch_related(
Prefetch("articles",
queryset=Article.objects.published().select_related("category"))
)
```
- `select_related` only follows single-valued relations (FK, O2O). Use it for the JOIN.
- `prefetch_related` handles multi-valued relations and runs a separate `IN (...)` query.
- `Prefetch` is the escape hatch when the related set needs filtering, ordering, or its own
`select_related` — otherwise you re-introduce an N+1 inside the prefetch.
## Proving query counts
```python
from django.test import TestCase
class FeedQueryCount(TestCase):
def test_feed(self):
with self.assertNumQueries(2):
[a.author.name for a in Article.objects.select_related("author")]
```
In a shell, eyeball SQL with `django.db.connection.queries` (needs `DEBUG=True`) or wrap a block
in `CaptureQueriesContext`.
## Fetch less
```python
Article.objects.only("id", "slug", "title") # SELECT just these columns
Article.objects.defer("body") # everything except the heavy one
Article.objects.values("id", "slug") # dicts, no model instances
Article.objects.values_list("id", flat=True) # a flat list of ids
```
- `.exists()` for existence; `.count()` only when you truly need the integer.
- Never `list(Model.objects.all())` then filter/slice in Python — push it into the QuerySet.
## Aggregate in the DB, not in Python
```python
from django.db.models import Count, Sum, Q
Author.objects.annotate(
published=Count("articles", filter=Q(articles__status="published"))
)
Order.objects.aggregate(total=Sum("amount"))
```
A Python loop summing `.amount` over a queryset is one row fetched per iteration plus the math;
`annotate`/`aggregate` does it in a single SQL pass.
## Bulk writes
```python
Article.objects.bulk_create(objs, batch_size=500) # one round-trip
Article.objects.bulk_update(objs, ["status"], batch_size=500)
Article.objects.filter(status="draft").update(status="published") # set-based, no Python loop
```
`.update()` is a single UPDATE and skips `save()`/signals — use it for set-based changes, and a
loop of `.save()` only when you need per-row signals/`save()` logic.
## Indexing through the ORM
```python
class Meta:
indexes = [
models.Index(fields=["tenant", "status"]),
models.Index(fields=["-published_at"], name="article_recent"),
]
```
Add the index where your hot `filter`/`order_by` actually hits. *Which* index, and reading
`EXPLAIN ANALYZE`, is `postgresdb`; here you just express it. Inspect a query plan from the ORM
with `qs.explain()`.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| finishing-a-development-branchobra/superpowersUse when implementation is complete, all tests pass, and you need to decide how to integrate the work | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| systematic-debuggingobra/superpowersUse when encountering any bug, test failure, or unexpected behavior, before proposing fixes | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| test-driven-developmentobra/superpowersUse when implementing any feature or bugfix, before writing implementation code | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| skill-creatoranthropics/skillsCreate new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, edit, or optimize an existing skill, run evals to test a skill, benchmark skill performance with variance analysis, or optimize a skill's description for better triggering accuracy. | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |