M2: cut the hosted product away from the local one
94 files, +1,395 -6,578. Three files are new; twenty-four are gone. The milestone is subtraction, and what is left is the single-user local storyteller the specification describes. Removed in full: campaign scripting and its QuickJS sandbox; multi-user accounts, guest sessions, login, registration and the shared demo key; the visitor-analytics tables, dashboard and page beacon; the access log of sign-ins, addresses and devices; per-IP and per-user rate limiting and quotas; Render deployment config; Postgres and psycopg; cloud inference providers, the API-key field and the key encryption that existed to store it; session-cookie signing. None of it was hidden behind a flag — the routes are gone and answer 404. Two things were kept that the brief allowed keeping. The `users` table and its foreign keys stay as an internal ownership detail, because rewriting them out means a migration across most of the schema to delete a column that costs nothing; nothing creates a second user and no request carries an identity. Five inert tables and four inert columns stay for the same reason, so an M1 campaign database opens unchanged. The one addition is app/endpoints.py, which decides where a story may be sent. Loopback, RFC1918, link-local, unique-local and CGNAT — an explicit allowlist of networks, not a guess at what `ipaddress` means by "private", which calls the documentation ranges private and IPv6 loopback reserved. Every address a hostname resolves to must be in it, so a split answer does not squeak through, and the rule runs both when the endpoint is saved and before every outbound request, because a name that resolved to the LAN this morning can resolve elsewhere this afternoon. Known cloud hosts are named in the refusal so the error says why rather than looking like broken DNS. TLS is never traded against it: M1's shared trust context is intact on all four clients and there is no way to skip verification. The hardcoded 120-second model timeout is now a setting. That was not theoretical — on this GPU-less four-core host a cold load of qwen2.5:3b-instruct took 648.9 seconds to produce the first turn, while turns 2 to 5 of the same campaign took 3.6 to 13.1. Connect stays short at 10s so a wrong address still fails fast; the read timeout defaults to 300s and is bounded at 3600, because "wait longer" must stay a number. Two defects found while testing and fixed here. An unknown /api path fell through the SPA catch-all and came back as HTML with status 200, so a client asking for JSON parsed a web page instead of learning the route was gone. And AIDND_CORS_ORIGINS accepted "*", which on an unauthenticated loopback API would hand every page on the Internet a write handle on the campaign database; it now refuses to start. Verified rather than assumed. Offline, on a network with no route out and no DNS: five turns, retry with both takes retained, restart with an identical transcript digest, a failed model call leaving the accepted AI-turn count untouched, and a capture with zero non-loopback unicast packets. Against a real second machine on the LAN over HTTPS with a private CA: four turns, restart, and a capture showing 289 packets to the approved host, 344 loopback, zero anywhere else, zero DNS queries. Cloud and public endpoints refused with their reasons; no API key settable; every removed route 404. 604 backend tests pass, down from 648 by the fifteen retired with the subsystems they tested and up by the twenty-nine added for the endpoint policy and the removed surface. The scripting tests were not deleted: eight files used a JavaScript counter as instrumentation for the state snapshot and rollback machinery, which M2 does not touch, so the counter moved to the world-state engine and those tests still assert what they always did. Frontend lint and build are clean; the image builds, and its wheel-building stage is gone with quickjs. No M3 work. Undo is still destructive and there is still no Redo.
This commit is contained in:
JesseMarkowitz
@@ -0,0 +1,181 @@
|
||||
"""What M2 removed stays removed, and what it must not break stays working.
|
||||
|
||||
A subtractive milestone needs tests that fail if the surface grows back. These
|
||||
are cheap, blunt, and deliberately not clever: they assert against the running
|
||||
app's route table, the shipped configuration files, and the settings API.
|
||||
|
||||
python -m pytest tests/test_local_only_surface.py -v
|
||||
"""
|
||||
import re
|
||||
from pathlib import Path
|
||||
|
||||
import pytest
|
||||
from fastapi import Depends
|
||||
from fastapi.testclient import TestClient
|
||||
|
||||
from app import auth, models
|
||||
from app.database import Base, SessionLocal, engine, get_db
|
||||
from app.main import app
|
||||
from app.providers.openai_compatible import (
|
||||
CONNECT_TIMEOUT, DEFAULT_READ_TIMEOUT, OpenAICompatibleProvider,
|
||||
)
|
||||
|
||||
REPO = Path(__file__).resolve().parents[2]
|
||||
|
||||
|
||||
def _paths() -> set[str]:
|
||||
"""Every path the app serves, including those inside included routers."""
|
||||
found = set()
|
||||
|
||||
def walk(routes):
|
||||
for r in routes:
|
||||
path = getattr(r, "path", None)
|
||||
if path:
|
||||
found.add(path)
|
||||
walk(getattr(r, "routes", []) or [])
|
||||
|
||||
walk(app.routes)
|
||||
return found
|
||||
|
||||
|
||||
@pytest.fixture()
|
||||
def client():
|
||||
Base.metadata.create_all(bind=engine)
|
||||
setup = SessionLocal()
|
||||
user = models.User(is_guest=False)
|
||||
setup.add(user)
|
||||
setup.flush()
|
||||
setup.add(models.Settings(user_id=user.id, model="test-model"))
|
||||
setup.commit()
|
||||
user_id = user.id
|
||||
setup.close()
|
||||
|
||||
def _current_user(db=Depends(get_db)):
|
||||
return db.get(models.User, user_id)
|
||||
|
||||
app.dependency_overrides[auth.get_current_user] = _current_user
|
||||
c = TestClient(app)
|
||||
try:
|
||||
yield c
|
||||
finally:
|
||||
app.dependency_overrides.clear()
|
||||
Base.metadata.drop_all(bind=engine)
|
||||
|
||||
|
||||
# --- the removed surfaces ------------------------------------------------
|
||||
|
||||
@pytest.mark.parametrize("prefix", ["/api/auth", "/api/analytics", "/api/scripts"])
|
||||
def test_no_route_serves_a_removed_subsystem(prefix):
|
||||
"""Accounts, the visitor dashboard, and campaign scripting are gone as
|
||||
routes, not merely hidden behind a flag."""
|
||||
assert not [p for p in _paths() if p.startswith(prefix)], prefix
|
||||
|
||||
|
||||
@pytest.mark.parametrize("path", [
|
||||
"/api/auth/me", "/api/auth/login", "/api/auth/register", "/api/auth/logout",
|
||||
"/api/analytics/summary", "/api/analytics/collect", "/api/analytics/access",
|
||||
"/api/scripts", "/api/adventures/1/scripts", "/api/adventures/1/script-state",
|
||||
])
|
||||
def test_a_removed_endpoint_answers_404(client, path):
|
||||
assert client.get(path).status_code == 404, path
|
||||
|
||||
|
||||
def test_the_application_has_no_scripting_engine():
|
||||
with pytest.raises(ImportError):
|
||||
__import__("app.scripting")
|
||||
|
||||
|
||||
def test_no_module_imports_quickjs():
|
||||
"""The dependency is gone from requirements; this catches an import that
|
||||
would put it back."""
|
||||
for py in (REPO / "backend" / "app").rglob("*.py"):
|
||||
assert "import quickjs" not in py.read_text(), py
|
||||
|
||||
|
||||
def test_requirements_carry_no_hosted_dependencies():
|
||||
text = (REPO / "backend" / "requirements.txt").read_text()
|
||||
for gone in ("quickjs", "psycopg", "cryptography"):
|
||||
assert gone not in text, gone
|
||||
|
||||
|
||||
def test_no_render_deployment_config():
|
||||
assert not (REPO / "render.yaml").exists()
|
||||
|
||||
|
||||
# --- no cloud provider, no key ------------------------------------------
|
||||
|
||||
def test_settings_expose_no_api_key_field(client):
|
||||
body = client.get("/api/settings").json()
|
||||
assert "api_key" not in body
|
||||
assert "has_api_key" not in body
|
||||
|
||||
|
||||
def test_an_api_key_cannot_be_set_through_the_api(client):
|
||||
"""Pydantic ignores unknown fields, so this asserts the value does not
|
||||
land rather than that the request is refused."""
|
||||
client.put("/api/settings", json={"api_key": "sk-should-not-stick"})
|
||||
db = SessionLocal()
|
||||
try:
|
||||
assert db.query(models.Settings).first().api_key == ""
|
||||
finally:
|
||||
db.close()
|
||||
|
||||
|
||||
def test_the_provider_sends_no_authorization_header():
|
||||
provider = OpenAICompatibleProvider("http://127.0.0.1:11434/v1", "m")
|
||||
assert "Authorization" not in provider._headers()
|
||||
|
||||
|
||||
# --- the model timeout ---------------------------------------------------
|
||||
|
||||
def test_the_default_timeout_is_generous_but_finite():
|
||||
"""M1 measured a cold model load exceeding the inherited hardcoded 120s on
|
||||
a CPU-only host. It must be longer than that, and it must be a number."""
|
||||
assert DEFAULT_READ_TIMEOUT > 120
|
||||
assert DEFAULT_READ_TIMEOUT <= 3600
|
||||
|
||||
|
||||
def test_connect_stays_short_while_reading_stays_patient():
|
||||
"""A wrong address should fail in seconds; a loading model should not."""
|
||||
provider = OpenAICompatibleProvider("http://127.0.0.1:11434/v1", "m")
|
||||
timeout = provider._timeout()
|
||||
assert timeout.connect == CONNECT_TIMEOUT <= 30
|
||||
assert timeout.read == DEFAULT_READ_TIMEOUT
|
||||
|
||||
|
||||
def test_the_timeout_is_configurable(client):
|
||||
r = client.put("/api/settings", json={"model_timeout_seconds": 900})
|
||||
assert r.status_code == 200, r.text
|
||||
assert client.get("/api/settings").json()["model_timeout_seconds"] == 900
|
||||
|
||||
|
||||
@pytest.mark.parametrize("value", [0, 29, 3601, -1])
|
||||
def test_an_unusable_timeout_is_refused(client, value):
|
||||
"""Not zero, not negative, and not "wait forever" spelled as a big number."""
|
||||
assert client.put(
|
||||
"/api/settings", json={"model_timeout_seconds": value}
|
||||
).status_code == 422
|
||||
|
||||
|
||||
def test_the_provider_honours_the_configured_timeout():
|
||||
provider = OpenAICompatibleProvider("http://127.0.0.1:11434/v1", "m", read_timeout=45)
|
||||
assert provider._timeout().read == 45
|
||||
|
||||
|
||||
# --- the storyteller stays on loopback -----------------------------------
|
||||
|
||||
def test_the_native_start_scripts_bind_loopback():
|
||||
for script in ("start.sh", "start.ps1"):
|
||||
text = (REPO / script).read_text(errors="ignore")
|
||||
assert "--host 127.0.0.1" in text, script
|
||||
assert "--host 0.0.0.0" not in text, script
|
||||
|
||||
|
||||
def test_compose_publishes_to_loopback_only():
|
||||
"""The container listens on 0.0.0.0 because a published port cannot reach
|
||||
anything else. What must stay loopback is the *host* side of the mapping."""
|
||||
text = (REPO / "docker-compose.yml").read_text()
|
||||
published = re.findall(r'^\s*-\s*"([^"]+)"', text, re.M)
|
||||
assert published, "no published ports found — has the file moved?"
|
||||
for mapping in published:
|
||||
assert mapping.startswith("127.0.0.1:"), mapping
|
||||
Reference in New Issue
Block a user