Backend · Claude Code (via permissions.deny)
Django 6 & Ninja / DRF .claudeignore
A .claudeignore for Django 6 & Ninja / DRF: the paths Claude Code should stay out of, with matching permissions.deny rules.
- Path
- .claudeignore
- Loads
- Pattern list
- Format
- gitignore syntax
- Used by
- Claude Code (via permissions.deny)
What this file tells the agent
Claude Code reads files while it works. This .claudeignore keeps the list of Django 6 & Ninja / DRF paths it should stay out of (build output, large binaries, logs and secrets). Copy the sensitive ones into permissions.deny in .claude/settings.json, which Claude Code enforces.
- Use asynchronous ORM operations (aget(), acreate(), afirst()) inside async def view handlers.
- Prevent N+1 query catastrophes: always apply select_related() for ForeignKey/OneToOne and prefetch_related() for ManyToMany/reverse relationships.
- Enforce Django Ninja schemas or DRF serializers with explicit field typing; never return raw unvalidated model instances.
- Never alter existing committed migration files; generate new sequential migrations via python manage.py makemigrations.
- Use GeneratedField for computed database columns rather than redundant model save() recalculations.
Code to avoid, and what to write instead
The rule steers the agent away from the left and toward the right.
# Discouraged: Synchronous blocking query inside async context & N+1 queries
@router.get("/articles/{article_id}")
async def get_article(request, article_id: int):
article = Article.objects.get(id=article_id) # Synchronous blocking call!
return {"id": article.id, "title": article.title, "author": article.author.username} # N+1 DB hit!from ninja import Router, Schema
from django.shortcuts import aget_object_or_404
from .models import Article
router = Router()
class ArticleOut(Schema):
id: int
title: str
author_name: str
@router.get("/articles/{article_id}", response=ArticleOut)
async def get_article(request, article_id: int):
# Asynchronous fetch with joined relationship preventing N+1
article = await Article.objects.select_related("author").aget(id=article_id)
return {
"id": article.id,
"title": article.title,
"author_name": article.author.username,
}Install it in one command
- 01
Go to your project root
The file goes at .claudeignore, relative to the root.
- 02
Download the file
Run the command on the right. It downloads this preset and writes the file.
- 03
Make sure it gets used
Claude Code does not enforce .claudeignore by itself. Copy the sensitive paths into permissions.deny in .claude/settings.json, which it does enforce.
$ curl -fsSL "https://www.devscratchpad.tech/api/raw/claudeignore/python-django" -o ".claudeignore"
writes .claudeignore
.claudeignore
To write the file from your own code instead, use Read your code or Convert Legacy Rules in the editor above.
More for Django 6 & Ninja / DRF
Questions
- Does Claude Code honor .claudeignore?
- Not by itself. Claude Code enforces permission rules, so copy the sensitive paths into permissions.deny in .claude/settings.json, for example Read(./.env). Keep .claudeignore in the repository root as the reviewable list.
- What should I put in .claudeignore?
- Exclude large build caches, minified bundles, lockfiles, and environment files containing sensitive credentials.