unsloth/studio/backend/tests/test_preview_followups.py
Daniel Han 80d3434d61
Studio: require signed capability tokens for /p preview links (#6666)
* Studio: require signed capability tokens for /p preview links

The public /p preview routes added in #6486 run model load and chat
generation as the admin user with no authentication. The only gate is the
preview ref, a deterministic outputs-root path (run or run/checkpoint) that
is guessable rather than secret. On a network-reachable Studio (--secure
tunnel or -H 0.0.0.0), an unauthenticated caller who guesses a ref can
consume GPU and probe a private fine-tuned checkpoint.

Make the share link an unguessable, revocable capability:

- Sign the canonical ref with a dedicated server-side secret (HMAC-SHA256,
  stored in app_secrets, independent of the JWT/login secret).
- Require a valid token on every /p chat, models, and page request before
  resolving a checkpoint or loading a model; missing or invalid tokens get a
  generic 404 so the surface never confirms a ref exists.
- Accept the token via ?k= (browser link and preview page) or
  Authorization: Bearer (OpenAI-compatible clients).
- Rotate the secret to revoke every outstanding link
  (POST /api/settings/preview-links/rotate).
- Clamp preview generation (max_tokens/max_completion_tokens <= 1024, n = 1)
  and set Referrer-Policy: no-referrer on the page so the token is not
  leaked via Referer.

Training history hands the authenticated owner the signed token, and the
copy-link button builds /p/{ref}?k={sig}.

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* Studio: honor a lower caller token limit in the preview clamp

Codex review: when only the legacy max_tokens was sent, the clamp left
max_completion_tokens at the 1024 default, and _effective_max_tokens prefers
max_completion_tokens, so a request like max_tokens=16 could still generate up
to 1024 tokens. Derive one effective limit (max_completion_tokens wins, else the
legacy max_tokens) and pin both fields to it so a caller's lower limit is kept.

* Studio: add preview kill switch, rate limit, and revoke-links UI

Follow-ups to the /p preview capability work:

- Public-sharing kill switch: a persisted setting (default on) gates the public
  /p surface. When off, every preview request 404s even with a valid token, and
  the owner UI stops offering share links. GET/PUT /api/settings/preview-sharing;
  enforced in _verify_or_404.
- Per-IP rate limit on the preview chat route: a coarse in-process sliding-window
  limiter (20 req/min/IP) returns 429 + Retry-After before the GPU lock is taken.
  Client IP honors X-Forwarded-For only when UNSLOTH_STUDIO_TRUST_FORWARDED is
  set, matching the login limiter's trust model.
- Settings UI: a "Preview sharing" section with the public-sharing toggle and a
  "Revoke all preview links" button (confirm dialog) that rotates the secret.

Tests cover the kill switch (404 when off), the 429 path, the sliding window,
client-IP trust behavior, and the setting default.

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* Studio: fix preview-fields sharing arg and refresh sigs after revoke

Codex review:
- P1: get_training_run_detail and update_training_run called _preview_fields
  with only output_dir after it gained a required sharing_on parameter, raising
  a 500 TypeError once get_run succeeded. Pass get_preview_sharing_enabled() at
  both sites; add a detail-endpoint regression test.
- P2: after rotating the preview secret from settings, the history grid still
  held stale preview_sig values, so a freshly copied link would 404. Emit
  emitTrainingRunsChanged() after a successful revoke so the grid refetches
  freshly signed refs.

* Studio: harden preview sharing controls (Codex review)

- Fail closed: a read failure on the preview-sharing kill switch now returns
  False instead of defaulting to enabled, so an unavailable settings DB can't
  reopen the public surface. A missing key still defaults to enabled.
- Per-IP rate limit behind the managed Cloudflare tunnel: client_ip now honors
  CF-Connecting-IP when the socket peer is loopback, so tunneled visitors are
  keyed by their real IP instead of collapsing onto the local cloudflared peer.
- GET /p no longer mints key/share_url when sharing is disabled; it returns
  sharing_enabled=false so clients don't distribute links that 404.
- Settings UI: toggling public sharing emits the training-runs-changed event so
  the history grid shows/hides Copy preview link without a manual refresh.

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* Studio: harden preview rate limiter and IP keying (Opus review)

From a two-agent review of the PR:

- Rate limiter no longer evicts an active bucket when the table is full: a flood
  of distinct keys could otherwise cycle out a throttled bucket and reset its
  counter. Evict only aged-out buckets; if the table is full of live clients,
  fail closed (deny the new key) instead.
- client_ip keys on the rightmost (proxy-appended) X-Forwarded-For hop when the
  trust env is set; the leftmost is client-spoofable. Documented the
  append/overwrite-proxy assumption.
- _verify_or_404 checks the capability token before the kill-switch DB read, so
  unauthenticated /p spam can't be used as an unbounded settings-DB sink and the
  response is identical regardless of the sharing on/off state.

Tests: nested run/checkpoint happy path + wrong-ref rejection, the eviction
fail-closed behavior, and route-level coverage for the rotate / preview-sharing
settings endpoints.

---------

Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
2026-06-25 21:40:48 -07:00

143 lines
5.5 KiB
Python

# SPDX-License-Identifier: AGPL-3.0-only
# Copyright 2026-present the Unsloth AI Inc. team. All rights reserved. See /studio/LICENSE.AGPL-3.0
"""Unit coverage for the preview follow-ups: rate limiter, client IP, kill switch."""
from pathlib import Path
import sys
import types as _types
_BACKEND_DIR = str(Path(__file__).resolve().parent.parent)
if _BACKEND_DIR not in sys.path:
sys.path.insert(0, _BACKEND_DIR)
_loggers_stub = _types.ModuleType("loggers")
_loggers_stub.get_logger = lambda name: __import__("logging").getLogger(name)
sys.modules.setdefault("loggers", _loggers_stub)
import utils.preview_rate_limit as rl
from utils.client_ip import client_ip
from utils.preview_sharing_settings import (
DEFAULT_PREVIEW_SHARING_ENABLED,
_coerce_bool,
get_preview_sharing_enabled,
)
# ── Rate limiter ─────────────────────────────────────────────────────────────
def test_rate_limit_per_key(monkeypatch):
monkeypatch.setattr(rl, "_MAX_REQUESTS", 3)
rl.reset()
assert rl.check_rate_limit("ip1") == 0
assert rl.check_rate_limit("ip1") == 0
assert rl.check_rate_limit("ip1") == 0
# 4th request over the ceiling -> positive retry-after seconds.
assert rl.check_rate_limit("ip1") > 0
# A different client is unaffected.
assert rl.check_rate_limit("ip2") == 0
def test_rate_limit_window_rolls_off(monkeypatch):
monkeypatch.setattr(rl, "_MAX_REQUESTS", 1)
monkeypatch.setattr(rl, "_WINDOW_SECONDS", 10.0)
rl.reset()
t = {"now": 1000.0}
monkeypatch.setattr(rl.time, "monotonic", lambda: t["now"])
assert rl.check_rate_limit("ip") == 0
assert rl.check_rate_limit("ip") > 0 # immediately over
t["now"] += 11.0 # window elapsed
assert rl.check_rate_limit("ip") == 0
def test_rate_limit_eviction_does_not_reset_active_bucket(monkeypatch):
# A flood of distinct keys must not cycle the table and clear a live limit.
monkeypatch.setattr(rl, "_MAX_REQUESTS", 1)
monkeypatch.setattr(rl, "_MAX_BUCKETS", 2)
rl.reset()
assert rl.check_rate_limit("a") == 0
assert rl.check_rate_limit("a") > 0 # 'a' throttled (active)
assert rl.check_rate_limit("b") == 0
assert rl.check_rate_limit("b") > 0 # 'b' throttled; table now full of actives
# A new key can't evict an active bucket -> denied (fail closed)...
assert rl.check_rate_limit("c") > 0
# ...and the flood did not reset 'a'.
assert rl.check_rate_limit("a") > 0
# ── Client IP ────────────────────────────────────────────────────────────────
class _Req:
def __init__(
self,
host,
headers = None,
):
self.client = _types.SimpleNamespace(host = host) if host else None
self.headers = headers or {}
def test_client_ip_uses_socket_peer_by_default(monkeypatch):
monkeypatch.delenv("UNSLOTH_STUDIO_TRUST_FORWARDED", raising = False)
# Forwarded header is ignored unless the operator opts in.
req = _Req("203.0.113.9", {"x-forwarded-for": "198.51.100.7"})
assert client_ip(req) == "203.0.113.9"
assert client_ip(None) == "_unknown"
def test_client_ip_uses_rightmost_forwarded_when_trusted(monkeypatch):
monkeypatch.setenv("UNSLOTH_STUDIO_TRUST_FORWARDED", "1")
# Leftmost is client-spoofable; the trusted proxy appends the real peer on the
# right, so the rightmost hop is the one we key on.
req = _Req("127.0.0.1", {"x-forwarded-for": "1.2.3.4, 198.51.100.7"})
assert client_ip(req) == "198.51.100.7"
def test_client_ip_uses_cf_connecting_ip_on_loopback(monkeypatch):
# Managed Cloudflare tunnel terminates at loopback; key by the real visitor.
monkeypatch.delenv("UNSLOTH_STUDIO_TRUST_FORWARDED", raising = False)
req = _Req("127.0.0.1", {"cf-connecting-ip": "198.51.100.7"})
assert client_ip(req) == "198.51.100.7"
def test_client_ip_ignores_cf_header_from_non_loopback(monkeypatch):
# A direct (non-loopback) caller can't spoof CF-Connecting-IP to skew the limit.
monkeypatch.delenv("UNSLOTH_STUDIO_TRUST_FORWARDED", raising = False)
req = _Req("203.0.113.9", {"cf-connecting-ip": "198.51.100.7"})
assert client_ip(req) == "203.0.113.9"
def test_client_ip_loopback_without_cf_returns_peer(monkeypatch):
monkeypatch.delenv("UNSLOTH_STUDIO_TRUST_FORWARDED", raising = False)
assert client_ip(_Req("127.0.0.1")) == "127.0.0.1"
# ── Kill-switch setting ──────────────────────────────────────────────────────
def test_sharing_defaults_enabled_and_coerces():
assert DEFAULT_PREVIEW_SHARING_ENABLED is True
assert _coerce_bool("off") is False
assert _coerce_bool("on") is True
assert _coerce_bool(True) is True
assert _coerce_bool("nonsense") is None
def test_sharing_missing_key_defaults_enabled(monkeypatch):
import storage.studio_db as sdb
monkeypatch.setattr(sdb, "get_app_setting", lambda key, fallback = None: None)
assert get_preview_sharing_enabled() is True
def test_sharing_read_error_fails_closed(monkeypatch):
# A transient settings-DB failure must not reopen the public surface.
import storage.studio_db as sdb
def _boom(*args, **kwargs):
raise RuntimeError("settings db unavailable")
monkeypatch.setattr(sdb, "get_app_setting", _boom)
assert get_preview_sharing_enabled() is False