Mode d’emploi
- Copiez la commande et examinez le paquet avant de l’installer dans votre environnement agent.
- Consultez le document original pour les outils, identifiants et environnements requis.
- Décrivez les entrées et le livrable souhaité ; comparez la sortie aux sources.
npx skills add https://github.com/sbknext/forge-frappe-skill --skill frappe-testingAvant de commencer
- L’installation n’inclut ni abonnement au modèle, ni quota API, ni accès aux fichiers.
- Étoiles et installations reflètent l’activité, pas un test indépendant de qualité ou de sécurité.
Rendu depuis le dépôt source en conservant titres, exemples, code, tableaux, liens et images.
Frappe Testing
Frappe ships a unittest-based harness with DB isolation. Tests run in a transaction that's rolled back, so they don't pollute the site.
Basic test
# myapp/myapp/doctype/widget/test_widget.py
import frappe
from frappe.tests.utils import FrappeTestCase
class TestWidget(FrappeTestCase):
def test_creates_with_defaults(self):
w = frappe.get_doc({"doctype": "Widget", "title": "A"}).insert()
self.assertEqual(w.status, "Open")
def test_validation_blocks_bad_qty(self):
with self.assertRaises(frappe.ValidationError):
frappe.get_doc({"doctype": "Widget", "title": "B", "qty": -1}).insert()FrappeTestCase wraps each test in a savepoint and rolls back — no manual cleanup needed.
Test records (fixtures)
# test_records.json next to the doctype → auto-loaded as dependencies
[ { "doctype": "Widget", "title": "Seed", "qty": 1 } ]Or build inline with a helper:
def make_widget(**kw):
return frappe.get_doc({"doctype": "Widget", "title": "T", **kw}).insert()Permission tests
def test_clerk_cannot_submit(self):
frappe.set_user("clerk@example.com")
self.assertFalse(frappe.has_permission("Widget", "submit"))
frappe.set_user("Administrator") # restoreMocking external calls
from unittest.mock import patch
@patch("myapp.services.sync.requests.post")
def test_sync_calls_api(self, mock_post):
mock_post.return_value.status_code = 200
run_sync("W-0001")
mock_post.assert_called_once()Running
bench --site <site> run-tests --app myapp
bench --site <site> run-tests --doctype "Widget"
bench --site <site> run-tests --module myapp.tests.test_sync
bench --site <site> run-tests --app myapp --coverage
bench --site <site> run-tests --skip-test-records --skip-before-tests # fast iteration
bench --site <site> run-tests --junit-xml-output=/tmp/junit.xml # CI outputEnable a test site: bench --site <site> set-config allow_tests true. Tests must run on a site whose name starts with test_ to prevent accidental data loss on production sites.
Test planning (manual + regression)
Before automating, define smoke (login + one submit flow) and critical-path cases in a structured format (preconditions → steps → expected). Cover permissions, hooks, API/import paths — not only Desk happy path. Template and checklists: skills/frappe/test-planner.md.
Rules
- Use
FrappeTestCase, not bareunittest.TestCase— you get DB rollback + helpers. - Never hit production data;
bench run-testsuses the site's test mode. - Make tests independent — don't rely on records another test created (savepoints roll back).
- For background jobs, call the method synchronously (
is_async=False) or invoke the function directly. - Assert behaviour and side-effects (DB rows, status), not implementation details.
Source: https://docs.frappe.io/framework/user/en/guides/automated-testing/unit-testing