Add configurable server security modes (#1773)

Fixes https://github.com/ArchiveBox/ArchiveBox/issues/239

## Summary
- add `SERVER_SECURITY_MODE` presets for safe subdomain replay, safe
one-domain no-JS replay, unsafe one-domain no-admin, and dangerous
one-domain full replay
- make host routing, replay URLs, static serving, and control-plane
access mode-aware
- add strict routing/header coverage plus a browser-backed
Chrome/Puppeteer test that verifies real same-origin behavior in all
four modes

## Testing
- `uv run pytest archivebox/tests/test_urls.py -v`
- `uv run pytest archivebox/tests/test_admin_views.py -v`
- `uv run pytest archivebox/tests/test_server_security_browser.py -v`

<!-- devin-review-badge-begin -->

---

<a href="https://app.devin.ai/review/archivebox/archivebox/pull/1773"
target="_blank">
  <picture>
<source media="(prefers-color-scheme: dark)"
srcset="https://static.devin.ai/assets/gh-open-in-devin-review-dark.svg?v=1">
<img
src="https://static.devin.ai/assets/gh-open-in-devin-review-light.svg?v=1"
alt="Open with Devin">
  </picture>
</a>
<!-- devin-review-badge-end -->


<!-- This is an auto-generated description by cubic. -->
---
## Summary by cubic
Adds configurable server security modes to isolate admin/API from
archived content, with a safe subdomain default and single-domain
fallbacks. Routing, replay endpoints, headers, and middleware are
mode-aware, with browser tests validating same-origin behavior.

- New Features
- Introduced SERVER_SECURITY_MODE with presets:
safe-subdomains-fullreplay (default), safe-onedomain-nojsreplay,
unsafe-onedomain-noadmin, danger-onedomain-fullreplay.
- Mode-aware routing and base URLs; one-domain modes use path-based
replay: /snapshot/<id>/... and /original/<domain>/....
- Control plane gate: block admin/API and non-GET methods in
unsafe-onedomain-noadmin; allow full access in
danger-onedomain-fullreplay.
- Safer replay: detect risky HTML/SVG and apply CSP sandbox (no scripts)
in safe-onedomain-nojsreplay; add X-ArchiveBox-Security-Mode and
X-Content-Type-Options: nosniff on replay responses.
- Middleware and serving: added ServerSecurityModeMiddleware, improved
HostRouting, and static server byte-range/CSP handling.
- Tests: added Chrome/Puppeteer browser tests and stricter URL routing
tests covering all modes.

- Migration
- Default requires wildcard subdomains for full isolation (admin., web.,
api., and snapshot-id.<base>).
- To run on one domain, set SERVER_SECURITY_MODE to a one-domain preset;
URLs switch to /snapshot/<id>/ and /original/<domain>/ paths.
- For production, prefer safe-subdomains-fullreplay; lower-security
modes print a startup warning.

<sup>Written for commit ad41b15581.
Summary will update on new commits.</sup>

<!-- End of auto-generated description by cubic. -->

archivebox/config/common.py

@@ -3,7 +3,7 @@ __package__ = "archivebox.config"
 import re
 import sys
 import shutil
-from typing import Dict, Optional, List
+from typing import ClassVar, Dict, Optional, List
 from pathlib import Path
 
 from rich import print
@@ -97,6 +97,13 @@ GENERAL_CONFIG = GeneralConfig()
 class ServerConfig(BaseConfigSet):
     toml_section_header: str = "SERVER_CONFIG"
 
+    SERVER_SECURITY_MODES: ClassVar[tuple[str, ...]] = (
+        "safe-subdomains-fullreplay",
+        "safe-onedomain-nojsreplay",
+        "unsafe-onedomain-noadmin",
+        "danger-onedomain-fullreplay",
+    )
+
     SECRET_KEY: str = Field(default_factory=lambda: get_random_string(50, "abcdefghijklmnopqrstuvwxyz0123456789_"))
     BIND_ADDR: str = Field(default="127.0.0.1:8000")
     LISTEN_HOST: str = Field(default="archivebox.localhost:8000")
@@ -104,6 +111,7 @@ class ServerConfig(BaseConfigSet):
     ARCHIVE_BASE_URL: str = Field(default="")
     ALLOWED_HOSTS: str = Field(default="*")
     CSRF_TRUSTED_ORIGINS: str = Field(default="http://admin.archivebox.localhost:8000")
+    SERVER_SECURITY_MODE: str = Field(default="safe-subdomains-fullreplay")
 
     SNAPSHOTS_PER_PAGE: int = Field(default=40)
     PREVIEW_ORIGINALS: bool = Field(default=True)
@@ -123,10 +131,89 @@ class ServerConfig(BaseConfigSet):
     REVERSE_PROXY_WHITELIST: str = Field(default="")
     LOGOUT_REDIRECT_URL: str = Field(default="/")
 
+    @field_validator("SERVER_SECURITY_MODE", mode="after")
+    def validate_server_security_mode(cls, v: str) -> str:
+        mode = (v or "").strip().lower()
+        if mode not in cls.SERVER_SECURITY_MODES:
+            raise ValueError(f"SERVER_SECURITY_MODE must be one of: {', '.join(cls.SERVER_SECURITY_MODES)}")
+        return mode
+
+    @property
+    def USES_SUBDOMAIN_ROUTING(self) -> bool:
+        return self.SERVER_SECURITY_MODE == "safe-subdomains-fullreplay"
+
+    @property
+    def ENABLES_FULL_JS_REPLAY(self) -> bool:
+        return self.SERVER_SECURITY_MODE in (
+            "safe-subdomains-fullreplay",
+            "unsafe-onedomain-noadmin",
+            "danger-onedomain-fullreplay",
+        )
+
+    @property
+    def CONTROL_PLANE_ENABLED(self) -> bool:
+        return self.SERVER_SECURITY_MODE != "unsafe-onedomain-noadmin"
+
+    @property
+    def BLOCK_UNSAFE_METHODS(self) -> bool:
+        return self.SERVER_SECURITY_MODE == "unsafe-onedomain-noadmin"
+
+    @property
+    def SHOULD_NEUTER_RISKY_REPLAY(self) -> bool:
+        return self.SERVER_SECURITY_MODE == "safe-onedomain-nojsreplay"
+
+    @property
+    def IS_UNSAFE_MODE(self) -> bool:
+        return self.SERVER_SECURITY_MODE == "unsafe-onedomain-noadmin"
+
+    @property
+    def IS_DANGEROUS_MODE(self) -> bool:
+        return self.SERVER_SECURITY_MODE == "danger-onedomain-fullreplay"
+
+    @property
+    def IS_LOWER_SECURITY_MODE(self) -> bool:
+        return self.SERVER_SECURITY_MODE in (
+            "unsafe-onedomain-noadmin",
+            "danger-onedomain-fullreplay",
+        )
+
 
 SERVER_CONFIG = ServerConfig()
 
 
+def _print_server_security_mode_warning() -> None:
+    if not SERVER_CONFIG.IS_LOWER_SECURITY_MODE:
+        return
+
+    print(
+        f"[yellow][!] WARNING: ArchiveBox is running with SERVER_SECURITY_MODE={SERVER_CONFIG.SERVER_SECURITY_MODE}[/yellow]",
+        file=sys.stderr,
+    )
+    print(
+        "[yellow]    Archived pages may share an origin with privileged app routes in this mode.[/yellow]",
+        file=sys.stderr,
+    )
+    print(
+        "[yellow]    To switch to the safer isolated setup:[/yellow]",
+        file=sys.stderr,
+    )
+    print(
+        "[yellow]    1. Set SERVER_SECURITY_MODE=safe-subdomains-fullreplay[/yellow]",
+        file=sys.stderr,
+    )
+    print(
+        "[yellow]    2. Point *.archivebox.localhost (or your chosen base domain) at this server[/yellow]",
+        file=sys.stderr,
+    )
+    print(
+        "[yellow]    3. Configure wildcard DNS/TLS or your reverse proxy so admin., web., api., and snapshot subdomains resolve[/yellow]",
+        file=sys.stderr,
+    )
+
+
+_print_server_security_mode_warning()
+
+
 class ArchivingConfig(BaseConfigSet):
     toml_section_header: str = "ARCHIVING_CONFIG"
 

archivebox/core/host_utils.py

@@ -55,6 +55,8 @@ def _build_listen_host(subdomain: str | None) -> str:
 
 
 def get_admin_host() -> str:
+    if not SERVER_CONFIG.USES_SUBDOMAIN_ROUTING:
+        return get_listen_host().lower()
     override = _normalize_base_url(SERVER_CONFIG.ADMIN_BASE_URL)
     if override:
         return urlparse(override).netloc.lower()
@@ -62,23 +64,33 @@ def get_admin_host() -> str:
 
 
 def get_web_host() -> str:
+    if not SERVER_CONFIG.USES_SUBDOMAIN_ROUTING:
+        return get_listen_host().lower()
     override = _normalize_base_url(SERVER_CONFIG.ARCHIVE_BASE_URL)
     if override:
         return urlparse(override).netloc.lower()
     return _build_listen_host("web")
 
 def get_api_host() -> str:
+    if not SERVER_CONFIG.USES_SUBDOMAIN_ROUTING:
+        return get_listen_host().lower()
     return _build_listen_host("api")
 
 def get_public_host() -> str:
+    if not SERVER_CONFIG.USES_SUBDOMAIN_ROUTING:
+        return get_listen_host().lower()
     return _build_listen_host("public")
 
 
 def get_snapshot_host(snapshot_id: str) -> str:
+    if not SERVER_CONFIG.USES_SUBDOMAIN_ROUTING:
+        return get_listen_host().lower()
     return _build_listen_host(snapshot_id)
 
 
 def get_original_host(domain: str) -> str:
+    if not SERVER_CONFIG.USES_SUBDOMAIN_ROUTING:
+        return get_listen_host().lower()
     return _build_listen_host(domain)
 
 
@@ -87,6 +99,8 @@ def is_snapshot_subdomain(subdomain: str) -> bool:
 
 
 def get_listen_subdomain(request_host: str) -> str:
+    if not SERVER_CONFIG.USES_SUBDOMAIN_ROUTING:
+        return ""
     req_host, req_port = split_host_port(request_host)
     listen_host, listen_port = get_listen_parts()
     if not listen_host:
@@ -127,6 +141,8 @@ def _build_base_url_for_host(host: str, request=None) -> str:
 
 
 def get_admin_base_url(request=None) -> str:
+    if not SERVER_CONFIG.USES_SUBDOMAIN_ROUTING:
+        return _build_base_url_for_host(get_listen_host(), request=request)
     override = _normalize_base_url(SERVER_CONFIG.ADMIN_BASE_URL)
     if override:
         return override
@@ -134,12 +150,16 @@ def get_admin_base_url(request=None) -> str:
 
 
 def get_web_base_url(request=None) -> str:
+    if not SERVER_CONFIG.USES_SUBDOMAIN_ROUTING:
+        return _build_base_url_for_host(get_listen_host(), request=request)
     override = _normalize_base_url(SERVER_CONFIG.ARCHIVE_BASE_URL)
     if override:
         return override
     return _build_base_url_for_host(get_web_host(), request=request)
 
 def get_api_base_url(request=None) -> str:
+    if not SERVER_CONFIG.USES_SUBDOMAIN_ROUTING:
+        return _build_base_url_for_host(get_listen_host(), request=request)
     return _build_base_url_for_host(get_api_host(), request=request)
 
 
@@ -149,10 +169,14 @@ def get_archive_base_url(request=None) -> str:
 
 
 def get_snapshot_base_url(snapshot_id: str, request=None) -> str:
+    if not SERVER_CONFIG.USES_SUBDOMAIN_ROUTING:
+        return _build_url(get_web_base_url(request=request), f"/snapshot/{snapshot_id}")
     return _build_base_url_for_host(get_snapshot_host(snapshot_id), request=request)
 
 
 def get_original_base_url(domain: str, request=None) -> str:
+    if not SERVER_CONFIG.USES_SUBDOMAIN_ROUTING:
+        return _build_url(get_web_base_url(request=request), f"/original/{domain}")
     return _build_base_url_for_host(get_original_host(domain), request=request)
 
 

archivebox/core/middleware.py

@@ -10,7 +10,7 @@ from django.core.exceptions import ImproperlyConfigured
 from django.shortcuts import redirect
 from django.contrib.staticfiles import finders
 from django.utils.http import http_date
-from django.http import HttpResponseNotModified
+from django.http import HttpResponseForbidden, HttpResponseNotModified
 
 from archivebox.config.common import SERVER_CONFIG
 from archivebox.config import VERSION
@@ -26,6 +26,7 @@ from archivebox.core.host_utils import (
     get_web_host,
     host_matches,
     is_snapshot_subdomain,
+    split_host_port,
 )
 from archivebox.core.views import SnapshotHostView, OriginalDomainHostView
 
@@ -90,6 +91,29 @@ def CacheControlMiddleware(get_response):
     return middleware
 
 
+def ServerSecurityModeMiddleware(get_response):
+    blocked_prefixes = ("/admin", "/accounts", "/api", "/add", "/web")
+    allowed_methods = {"GET", "HEAD", "OPTIONS"}
+
+    def middleware(request):
+        if SERVER_CONFIG.CONTROL_PLANE_ENABLED:
+            return get_response(request)
+
+        request.user = AnonymousUser()
+        request._cached_user = request.user
+
+        if request.method.upper() not in allowed_methods:
+            return HttpResponseForbidden("ArchiveBox is running with the control plane disabled in this security mode.")
+
+        for prefix in blocked_prefixes:
+            if request.path == prefix or request.path.startswith(f"{prefix}/"):
+                return HttpResponseForbidden("ArchiveBox is running with the control plane disabled in this security mode.")
+
+        return get_response(request)
+
+    return middleware
+
+
 def HostRoutingMiddleware(get_response):
     def middleware(request):
         request_host = (request.get_host() or "").lower()
@@ -100,6 +124,21 @@ def HostRoutingMiddleware(get_response):
         listen_host = get_listen_host()
         subdomain = get_listen_subdomain(request_host)
 
+        if not SERVER_CONFIG.USES_SUBDOMAIN_ROUTING:
+            if host_matches(request_host, listen_host):
+                return get_response(request)
+
+            req_host, req_port = split_host_port(request_host)
+            listen_host_only, listen_port = split_host_port(listen_host)
+            if req_host.endswith(f".{listen_host_only}"):
+                if not listen_port or not req_port or listen_port == req_port:
+                    target = build_web_url(request.path, request=request)
+                    if request.META.get("QUERY_STRING"):
+                        target = f"{target}?{request.META['QUERY_STRING']}"
+                    return redirect(target)
+
+            return get_response(request)
+
         if host_matches(request_host, admin_host):
             return get_response(request)
 

archivebox/core/settings.py

@@ -86,6 +86,7 @@ MIDDLEWARE = [
     "django.middleware.csrf.CsrfViewMiddleware",
     "django.contrib.auth.middleware.AuthenticationMiddleware",
     "archivebox.core.middleware.ReverseProxyAuthMiddleware",
+    "archivebox.core.middleware.ServerSecurityModeMiddleware",
     "archivebox.core.middleware.HostRoutingMiddleware",
     "django.contrib.messages.middleware.MessageMiddleware",
     "archivebox.core.middleware.CacheControlMiddleware",

archivebox/core/urls.py

@@ -9,7 +9,7 @@ from django.http import HttpRequest
 from archivebox.misc.serve_static import serve_static
 
 from archivebox.core.admin_site import archivebox_admin
-from archivebox.core.views import HomepageView, SnapshotView, SnapshotPathView, PublicIndexView, AddView, WebAddView, HealthCheckView, live_progress_view
+from archivebox.core.views import HomepageView, SnapshotView, SnapshotPathView, SnapshotReplayView, OriginalDomainReplayView, PublicIndexView, AddView, WebAddView, HealthCheckView, live_progress_view
 
 
 # GLOBAL_CONTEXT doesn't work as-is, disabled for now: https://github.com/ArchiveBox/ArchiveBox/discussions/1306
@@ -33,6 +33,8 @@ urlpatterns = [
     
     path('archive/', RedirectView.as_view(url='/')),
     path('archive/<path:path>', SnapshotView.as_view(), name='Snapshot'),
+    re_path(r'^snapshot\/(?P<snapshot_id>[0-9a-fA-F-]{8,36})(?:\/(?P<path>.*))?$', SnapshotReplayView.as_view(), name='snapshot-replay'),
+    re_path(r'^original\/(?P<domain>[^/]+)(?:\/(?P<path>.*))?$', OriginalDomainReplayView.as_view(), name='original-replay'),
     re_path(r'^web/(?P<url>(?!\d{4}(?:\d{2})?(?:\d{2})?(?:/|$)).+)$', WebAddView.as_view(), name='web-add'),
     re_path(r'^(?P<username>[^/]+)/(?P<date>\d{4}(?:\d{2})?(?:\d{2})?)/(?P<url>https?://.*)$', SnapshotPathView.as_view(), name='snapshot-path-url'),
     re_path(r'^(?P<username>[^/]+)/(?P<date>\d{4}(?:\d{2})?(?:\d{2})?)/(?P<domain>[^/]+)(?:/(?P<snapshot_id>[0-9a-fA-F-]{8,36})(?:/(?P<path>.*))?)?$', SnapshotPathView.as_view(), name='snapshot-path'),

archivebox/core/views.py

@@ -52,15 +52,21 @@ def _files_index_target(snapshot: Snapshot, archivefile: str | None) -> str:
     return target
 
 
+def _admin_login_redirect_or_forbidden(request: HttpRequest):
+    if SERVER_CONFIG.CONTROL_PLANE_ENABLED:
+        return redirect(f'/admin/login/?next={request.path}')
+    return HttpResponseForbidden("ArchiveBox is running with the control plane disabled in this security mode.")
+
+
 class HomepageView(View):
     def get(self, request):
-        if request.user.is_authenticated:
+        if request.user.is_authenticated and SERVER_CONFIG.CONTROL_PLANE_ENABLED:
             return redirect('/admin/core/snapshot/')
 
         if SERVER_CONFIG.PUBLIC_INDEX:
             return redirect('/public')
 
-        return redirect(f'/admin/login/?next={request.path}')
+        return _admin_login_redirect_or_forbidden(request)
 
 
 class SnapshotView(View):
@@ -277,7 +283,7 @@ class SnapshotView(View):
 
     def get(self, request, path):
         if not request.user.is_authenticated and not SERVER_CONFIG.PUBLIC_SNAPSHOTS:
-            return redirect(f'/admin/login/?next={request.path}')
+            return _admin_login_redirect_or_forbidden(request)
 
         snapshot = None
 
@@ -308,7 +314,7 @@ class SnapshotView(View):
                     if request.GET.get('files'):
                         target_path = _files_index_target(snapshot, archivefile)
                         response = serve_static_with_byterange_support(
-                            request, target_path, document_root=snapshot.output_dir, show_indexes=True,
+                            request, target_path, document_root=snapshot.output_dir, show_indexes=True, is_archive_replay=True,
                         )
                     elif archivefile == 'index.html':
                         # if they requested snapshot index, serve live rendered template instead of static html
@@ -474,7 +480,7 @@ class SnapshotPathView(View):
 
     def get(self, request, username: str, date: str | None = None, domain: str | None = None, snapshot_id: str | None = None, path: str = "", url: str | None = None):
         if not request.user.is_authenticated and not SERVER_CONFIG.PUBLIC_SNAPSHOTS:
-            return redirect(f'/admin/login/?next={request.path}')
+            return _admin_login_redirect_or_forbidden(request)
 
         if username == 'system':
             return redirect(request.path.replace('/system/', '/web/', 1))
@@ -573,14 +579,14 @@ class SnapshotPathView(View):
         if request.GET.get('files'):
             target_path = _files_index_target(snapshot, archivefile)
             return serve_static_with_byterange_support(
-                request, target_path, document_root=snapshot.output_dir, show_indexes=True,
+                request, target_path, document_root=snapshot.output_dir, show_indexes=True, is_archive_replay=True,
             )
 
         if archivefile == "index.html":
             return SnapshotView.render_live_index(request, snapshot)
 
         return serve_static_with_byterange_support(
-            request, archivefile, document_root=snapshot.output_dir, show_indexes=True,
+            request, archivefile, document_root=snapshot.output_dir, show_indexes=True, is_archive_replay=True,
         )
 
 
@@ -670,6 +676,7 @@ def _serve_responses_path(request, responses_root: Path, rel_path: str, show_ind
                 candidate,
                 document_root=str(responses_root),
                 show_indexes=show_indexes,
+                is_archive_replay=True,
             )
         except Http404:
             pass
@@ -682,18 +689,85 @@ def _serve_responses_path(request, responses_root: Path, rel_path: str, show_ind
                 rel_dir,
                 document_root=str(responses_root),
                 show_indexes=True,
+                is_archive_replay=True,
             )
         except Http404:
             return None
     return None
 
 
+def _serve_snapshot_replay(request: HttpRequest, snapshot: Snapshot, path: str = ""):
+    rel_path = path or ""
+    show_indexes = bool(request.GET.get("files"))
+    if not rel_path or rel_path.endswith("/"):
+        if show_indexes:
+            rel_path = rel_path.rstrip("/")
+        else:
+            rel_path = f"{rel_path}index.html"
+    rel_path = _safe_archive_relpath(rel_path)
+    if rel_path is None:
+        raise Http404
+
+    try:
+        return serve_static_with_byterange_support(
+            request,
+            rel_path,
+            document_root=snapshot.output_dir,
+            show_indexes=show_indexes,
+            is_archive_replay=True,
+        )
+    except Http404:
+        pass
+
+    host = urlparse(snapshot.url).hostname or snapshot.domain
+    responses_root = Path(snapshot.output_dir) / "responses" / host
+    if responses_root.exists():
+        response = _serve_responses_path(request, responses_root, rel_path, show_indexes)
+        if response is not None:
+            return response
+
+    raise Http404
+
+
+def _serve_original_domain_replay(request: HttpRequest, domain: str, path: str = ""):
+    rel_path = path or ""
+    if not rel_path or rel_path.endswith("/"):
+        rel_path = f"{rel_path}index.html"
+    rel_path = _safe_archive_relpath(rel_path)
+    if rel_path is None:
+        raise Http404
+
+    domain = domain.lower()
+    match = _latest_response_match(domain, rel_path)
+    if not match and "." not in Path(rel_path).name:
+        index_path = f"{rel_path.rstrip('/')}/index.html"
+        match = _latest_response_match(domain, index_path)
+    if not match and "." not in Path(rel_path).name:
+        html_path = f"{rel_path}.html"
+        match = _latest_response_match(domain, html_path)
+
+    show_indexes = bool(request.GET.get("files"))
+    if match:
+        responses_root, rel_to_root = match
+        response = _serve_responses_path(request, responses_root, str(rel_to_root), show_indexes)
+        if response is not None:
+            return response
+
+    responses_root = _latest_responses_root(domain)
+    if responses_root:
+        response = _serve_responses_path(request, responses_root, rel_path, show_indexes)
+        if response is not None:
+            return response
+
+    raise Http404
+
+
 class SnapshotHostView(View):
     """Serve snapshot directory contents on <snapshot_id>.<listen_host>/<path>."""
 
     def get(self, request, snapshot_id: str, path: str = ""):
         if not request.user.is_authenticated and not SERVER_CONFIG.PUBLIC_SNAPSHOTS:
-            return HttpResponseForbidden("Public snapshots are disabled.")
+            return _admin_login_redirect_or_forbidden(request)
         snapshot = None
         if snapshot_id:
             try:
@@ -708,37 +782,30 @@ class SnapshotHostView(View):
 
         if not snapshot:
             raise Http404
+        return _serve_snapshot_replay(request, snapshot, path)
 
-        rel_path = path or ""
-        show_indexes = bool(request.GET.get("files"))
-        if not rel_path or rel_path.endswith("/"):
-            if show_indexes:
-                rel_path = rel_path.rstrip("/")
-            else:
-                rel_path = f"{rel_path}index.html"
-        rel_path = _safe_archive_relpath(rel_path)
-        if rel_path is None:
-            raise Http404
+
+class SnapshotReplayView(View):
+    """Serve snapshot directory contents on a one-domain replay path."""
+
+    def get(self, request, snapshot_id: str, path: str = ""):
+        if not request.user.is_authenticated and not SERVER_CONFIG.PUBLIC_SNAPSHOTS:
+            return _admin_login_redirect_or_forbidden(request)
 
         try:
-            return serve_static_with_byterange_support(
-                request,
-                rel_path,
-                document_root=snapshot.output_dir,
-                show_indexes=show_indexes,
-            )
-        except Http404:
-            pass
+            snapshot = Snapshot.objects.get(pk=snapshot_id)
+        except Snapshot.DoesNotExist:
+            try:
+                snapshot = Snapshot.objects.get(id__startswith=snapshot_id)
+            except Snapshot.DoesNotExist:
+                raise Http404
+            except Snapshot.MultipleObjectsReturned:
+                snapshot = Snapshot.objects.filter(id__startswith=snapshot_id).first()
 
-        # Fallback to responses/<domain>/<path>
-        host = urlparse(snapshot.url).hostname or snapshot.domain
-        responses_root = Path(snapshot.output_dir) / "responses" / host
-        if responses_root.exists():
-            response = _serve_responses_path(request, responses_root, rel_path, show_indexes)
-            if response is not None:
-                return response
+        if snapshot is None:
+            raise Http404
 
-        raise Http404
+        return _serve_snapshot_replay(request, snapshot, path)
 
 
 class OriginalDomainHostView(View):
@@ -746,38 +813,17 @@ class OriginalDomainHostView(View):
 
     def get(self, request, domain: str, path: str = ""):
         if not request.user.is_authenticated and not SERVER_CONFIG.PUBLIC_SNAPSHOTS:
-            return HttpResponseForbidden("Public snapshots are disabled.")
-        rel_path = path or ""
-        if not rel_path or rel_path.endswith("/"):
-            rel_path = f"{rel_path}index.html"
-        rel_path = _safe_archive_relpath(rel_path)
-        if rel_path is None:
-            raise Http404
+            return _admin_login_redirect_or_forbidden(request)
+        return _serve_original_domain_replay(request, domain, path)
 
-        domain = domain.lower()
-        match = _latest_response_match(domain, rel_path)
-        if not match and "." not in Path(rel_path).name:
-            index_path = f"{rel_path.rstrip('/')}/index.html"
-            match = _latest_response_match(domain, index_path)
-        if not match and "." not in Path(rel_path).name:
-            html_path = f"{rel_path}.html"
-            match = _latest_response_match(domain, html_path)
 
-        show_indexes = bool(request.GET.get("files"))
-        if match:
-            responses_root, rel_to_root = match
-            response = _serve_responses_path(request, responses_root, str(rel_to_root), show_indexes)
-            if response is not None:
-                return response
+class OriginalDomainReplayView(View):
+    """Serve original-domain replay content on a one-domain replay path."""
 
-        # If no direct match, try serving directory index from latest responses root
-        responses_root = _latest_responses_root(domain)
-        if responses_root:
-            response = _serve_responses_path(request, responses_root, rel_path, show_indexes)
-            if response is not None:
-                return response
-
-        raise Http404
+    def get(self, request, domain: str, path: str = ""):
+        if not request.user.is_authenticated and not SERVER_CONFIG.PUBLIC_SNAPSHOTS:
+            return _admin_login_redirect_or_forbidden(request)
+        return _serve_original_domain_replay(request, domain, path)
 
 
 class PublicIndexView(ListView):
@@ -834,7 +880,7 @@ class PublicIndexView(ListView):
             response = super().get(*args, **kwargs)
             return response
         else:
-            return redirect(f'/admin/login/?next={self.request.path}')
+            return _admin_login_redirect_or_forbidden(self.request)
 
 @method_decorator(csrf_exempt, name='dispatch')
 class AddView(UserPassesTestMixin, FormView):

archivebox/misc/serve_static.py

@@ -81,6 +81,17 @@ MARKDOWN_BOLD_RE = re.compile(r'\*\*([^*]+)\*\*')
 MARKDOWN_ITALIC_RE = re.compile(r'(?<!\*)\*([^*]+)\*(?!\*)')
 HTML_TAG_RE = re.compile(r'<[A-Za-z][^>]*>')
 HTML_BODY_RE = re.compile(r'<body[^>]*>(.*)</body>', flags=re.IGNORECASE | re.DOTALL)
+RISKY_REPLAY_MIMETYPES = {
+    "text/html",
+    "application/xhtml+xml",
+    "image/svg+xml",
+}
+RISKY_REPLAY_EXTENSIONS = {".html", ".htm", ".xhtml", ".svg", ".svgz"}
+RISKY_REPLAY_MARKERS = (
+    "<!doctype html",
+    "<html",
+    "<svg",
+)
 
 
 def _extract_markdown_candidate(text: str) -> str:
@@ -278,7 +289,56 @@ def _render_markdown_document(markdown_text: str) -> str:
     return wrapped
 
 
-def serve_static_with_byterange_support(request, path, document_root=None, show_indexes=False):
+def _content_type_base(content_type: str) -> str:
+    return (content_type or "").split(";", 1)[0].strip().lower()
+
+
+def _is_risky_replay_document(fullpath: Path, content_type: str) -> bool:
+    if fullpath.suffix.lower() in RISKY_REPLAY_EXTENSIONS:
+        return True
+
+    if _content_type_base(content_type) in RISKY_REPLAY_MIMETYPES:
+        return True
+
+    # Unknown archived response paths often have no extension. Sniff a small prefix
+    # so one-domain no-JS mode still catches HTML/SVG documents.
+    try:
+        head = fullpath.read_bytes()[:4096].decode("utf-8", errors="ignore").lower()
+    except Exception:
+        return False
+
+    return any(marker in head for marker in RISKY_REPLAY_MARKERS)
+
+
+def _apply_archive_replay_headers(response: HttpResponse, *, fullpath: Path, content_type: str, is_archive_replay: bool) -> HttpResponse:
+    if not is_archive_replay:
+        return response
+
+    response.headers.setdefault("X-Content-Type-Options", "nosniff")
+    response.headers.setdefault("X-ArchiveBox-Security-Mode", SERVER_CONFIG.SERVER_SECURITY_MODE)
+
+    if SERVER_CONFIG.SHOULD_NEUTER_RISKY_REPLAY and _is_risky_replay_document(fullpath, content_type):
+        response.headers["Content-Security-Policy"] = (
+            "sandbox; "
+            "default-src 'self' data: blob:; "
+            "script-src 'none'; "
+            "object-src 'none'; "
+            "base-uri 'none'; "
+            "form-action 'none'; "
+            "connect-src 'none'; "
+            "worker-src 'none'; "
+            "frame-ancestors 'self'; "
+            "style-src 'self' 'unsafe-inline' data: blob:; "
+            "img-src 'self' data: blob:; "
+            "media-src 'self' data: blob:; "
+            "font-src 'self' data: blob:;"
+        )
+        response.headers.setdefault("Referrer-Policy", "no-referrer")
+
+    return response
+
+
+def serve_static_with_byterange_support(request, path, document_root=None, show_indexes=False, is_archive_replay: bool=False):
     """
     Overrides Django's built-in django.views.static.serve function to support byte range requests.
     This allows you to do things like seek into the middle of a huge mp4 or WACZ without downloading the whole file.
@@ -289,7 +349,8 @@ def serve_static_with_byterange_support(request, path, document_root=None, show_
     fullpath = Path(safe_join(document_root, path))
     if os.access(fullpath, os.R_OK) and fullpath.is_dir():
         if show_indexes:
-            return static.directory_index(path, fullpath)
+            response = static.directory_index(path, fullpath)
+            return _apply_archive_replay_headers(response, fullpath=fullpath, content_type="text/html", is_archive_replay=is_archive_replay)
         raise Http404(_("Directory indexes are not allowed here."))
     if not os.access(fullpath, os.R_OK):
         raise Http404(_("“%(path)s” does not exist") % {"path": fullpath})
@@ -312,7 +373,7 @@ def serve_static_with_byterange_support(request, path, document_root=None, show_
                 not_modified.headers["ETag"] = etag
                 not_modified.headers["Cache-Control"] = f"{_cache_policy()}, max-age=31536000, immutable"
                 not_modified.headers["Last-Modified"] = http_date(statobj.st_mtime)
-                return not_modified
+                return _apply_archive_replay_headers(not_modified, fullpath=fullpath, content_type="", is_archive_replay=is_archive_replay)
     
     content_type, encoding = mimetypes.guess_type(str(fullpath))
     content_type = content_type or "application/octet-stream"
@@ -333,7 +394,7 @@ def serve_static_with_byterange_support(request, path, document_root=None, show_
     # Respect the If-Modified-Since header for non-markdown responses.
     if not (content_type.startswith("text/plain") or content_type.startswith("text/html")):
         if not static.was_modified_since(request.META.get("HTTP_IF_MODIFIED_SINCE"), statobj.st_mtime):
-            return HttpResponseNotModified()
+            return _apply_archive_replay_headers(HttpResponseNotModified(), fullpath=fullpath, content_type=content_type, is_archive_replay=is_archive_replay)
 
     # Heuristic fix: some archived HTML outputs (e.g. mercury content.html)
     # are stored with HTML-escaped markup or markdown sources. If so, render sensibly.
@@ -360,7 +421,7 @@ def serve_static_with_byterange_support(request, path, document_root=None, show_
                     response.headers["Content-Disposition"] = f'inline; filename="{fullpath.name}"'
                     if encoding:
                         response.headers["Content-Encoding"] = encoding
-                    return response
+                    return _apply_archive_replay_headers(response, fullpath=fullpath, content_type="text/html; charset=utf-8", is_archive_replay=is_archive_replay)
                 if escaped_count and escaped_count > tag_count * 2:
                     response = HttpResponse(decoded, content_type=content_type)
                     response.headers["Last-Modified"] = http_date(statobj.st_mtime)
@@ -372,7 +433,7 @@ def serve_static_with_byterange_support(request, path, document_root=None, show_
                     response.headers["Content-Disposition"] = f'inline; filename="{fullpath.name}"'
                     if encoding:
                         response.headers["Content-Encoding"] = encoding
-                    return response
+                    return _apply_archive_replay_headers(response, fullpath=fullpath, content_type=content_type, is_archive_replay=is_archive_replay)
         except Exception:
             pass
 
@@ -416,7 +477,7 @@ def serve_static_with_byterange_support(request, path, document_root=None, show_
                 response.status_code = 206
     if encoding:
         response.headers["Content-Encoding"] = encoding
-    return response
+    return _apply_archive_replay_headers(response, fullpath=fullpath, content_type=content_type, is_archive_replay=is_archive_replay)
 
 
 def serve_static(request, path, **kwargs):

archivebox/tests/test_server_security_browser.py

@@ -0,0 +1,578 @@
+#!/usr/bin/env python3
+"""Browser-level security mode tests using the existing Node/Puppeteer runtime."""
+
+from __future__ import annotations
+
+import json
+import os
+import shutil
+import signal
+import socket
+import subprocess
+import sys
+import textwrap
+import time
+from pathlib import Path
+from urllib.parse import urlencode
+
+import pytest
+import requests
+
+from .conftest import _ensure_puppeteer, _find_cached_chromium, _find_system_browser, run_python_cwd
+
+
+PUPPETEER_PROBE_SCRIPT = """\
+const fs = require("node:fs");
+const puppeteer = require("puppeteer");
+
+async function login(page, config) {
+  const result = {
+    reachable: false,
+    succeeded: false,
+    finalUrl: null,
+    status: null,
+    error: null,
+  };
+
+  try {
+    const response = await page.goto(config.adminLoginUrl, {
+      waitUntil: "networkidle2",
+      timeout: 15000,
+    });
+    result.reachable = true;
+    result.status = response ? response.status() : null;
+
+    const usernameInput = await page.$('input[name="username"]');
+    const passwordInput = await page.$('input[name="password"]');
+    if (!usernameInput || !passwordInput) {
+      result.finalUrl = page.url();
+      return result;
+    }
+
+    await usernameInput.type(config.username);
+    await passwordInput.type(config.password);
+    await Promise.all([
+      page.waitForNavigation({waitUntil: "networkidle2", timeout: 15000}),
+      page.click('button[type="submit"], input[type="submit"]'),
+    ]);
+
+    result.finalUrl = page.url();
+    result.succeeded = !page.url().includes("/admin/login/");
+    return result;
+  } catch (error) {
+    result.error = String(error);
+    result.finalUrl = page.url();
+    return result;
+  }
+}
+
+async function main() {
+  const config = JSON.parse(fs.readFileSync(0, "utf8"));
+  const browser = await puppeteer.launch({
+    executablePath: config.chromePath,
+    headless: true,
+    args: [
+      "--no-sandbox",
+      "--disable-dev-shm-usage",
+      "--disable-background-networking",
+    ],
+  });
+
+  const loginPage = await browser.newPage();
+  const loginResult = await login(loginPage, config);
+  await loginPage.close();
+
+  const page = await browser.newPage();
+  const consoleMessages = [];
+  const requestFailures = [];
+  page.on("console", (message) => {
+    consoleMessages.push({type: message.type(), text: message.text()});
+  });
+  page.on("pageerror", (error) => {
+    consoleMessages.push({type: "pageerror", text: String(error)});
+  });
+  page.on("requestfailed", (request) => {
+    requestFailures.push({
+      url: request.url(),
+      error: request.failure() ? request.failure().errorText : "unknown",
+    });
+  });
+
+  const response = await page.goto(config.dangerousUrl, {
+    waitUntil: "networkidle2",
+    timeout: 15000,
+  });
+
+  await new Promise((resolve) => setTimeout(resolve, 1500));
+
+  const pageState = await page.evaluate(() => ({
+    href: location.href,
+    scriptRan: window.__dangerousScriptRan === true,
+    probeResults: window.__probeResults || null,
+    bodyText: document.body ? document.body.innerText.slice(0, 600) : "",
+  }));
+
+  const output = {
+    mode: config.mode,
+    login: loginResult,
+    dangerousPage: {
+      status: response ? response.status() : null,
+      finalUrl: page.url(),
+      contentSecurityPolicy: response ? response.headers()["content-security-policy"] || null : null,
+      archiveboxSecurityMode: response ? response.headers()["x-archivebox-security-mode"] || null : null,
+    },
+    pageState,
+    consoleMessages,
+    requestFailures,
+  };
+
+  console.log(JSON.stringify(output));
+  await browser.close();
+}
+
+main().catch((error) => {
+  console.error(String(error));
+  process.exit(1);
+});
+"""
+
+
+def _resolve_browser(shared_lib: Path) -> Path | None:
+    env_browser = os.environ.get("CHROME_BINARY") or os.environ.get("CHROME_BIN")
+    if env_browser:
+        candidate = Path(env_browser).expanduser()
+        if candidate.exists():
+            return candidate
+
+    cached = _find_cached_chromium(shared_lib)
+    if cached and cached.exists():
+        return cached
+
+    system = _find_system_browser()
+    if system and system.exists():
+        return system
+
+    which_candidates = ("chromium", "chromium-browser", "google-chrome", "google-chrome-stable", "chrome")
+    for binary in which_candidates:
+        resolved = shutil.which(binary)
+        if resolved:
+            return Path(resolved)
+
+    mac_candidates = (
+        Path("/Applications/Google Chrome.app/Contents/MacOS/Google Chrome"),
+        Path("/Applications/Chromium.app/Contents/MacOS/Chromium"),
+    )
+    for candidate in mac_candidates:
+        if candidate.exists():
+            return candidate
+
+    return None
+
+
+@pytest.fixture(scope="session")
+def browser_runtime(tmp_path_factory):
+    if shutil.which("node") is None or shutil.which("npm") is None:
+        pytest.skip("Node.js and npm are required for browser security tests")
+
+    shared_lib = tmp_path_factory.mktemp("archivebox_browser_lib")
+    _ensure_puppeteer(shared_lib)
+
+    browser = _resolve_browser(shared_lib)
+    if not browser:
+        pytest.skip("No Chrome/Chromium binary available for browser security tests")
+
+    return {
+        "node_modules_dir": shared_lib / "npm" / "node_modules",
+        "chrome_binary": browser,
+    }
+
+
+def _seed_archive(data_dir: Path) -> dict[str, object]:
+    script = textwrap.dedent(
+        """
+        import json
+        import os
+        from pathlib import Path
+        from django.utils import timezone
+
+        os.environ.setdefault("DJANGO_SETTINGS_MODULE", "archivebox.core.settings")
+        import django
+        django.setup()
+
+        from django.contrib.auth import get_user_model
+        from archivebox.core.models import Snapshot
+        from archivebox.crawls.models import Crawl
+
+        User = get_user_model()
+        admin, _ = User.objects.get_or_create(
+            username="testadmin",
+            defaults={"email": "admin@example.com", "is_staff": True, "is_superuser": True},
+        )
+        admin.set_password("testpassword")
+        admin.save()
+
+        snapshots = {}
+        fixture_specs = (
+            ("attacker", "https://attacker.example/entry", "Attacker Snapshot", "ATTACKER_SECRET"),
+            ("victim", "https://victim.example/private", "Victim Snapshot", "VICTIM_SECRET"),
+        )
+
+        for slug, url, title, secret in fixture_specs:
+            crawl = Crawl.objects.create(
+                urls=url,
+                created_by=admin,
+                status=Crawl.StatusChoices.SEALED,
+                retry_at=timezone.now(),
+            )
+            snapshot = Snapshot.objects.create(
+                url=url,
+                title=title,
+                crawl=crawl,
+                status=Snapshot.StatusChoices.SEALED,
+                downloaded_at=timezone.now(),
+            )
+            output_dir = Path(snapshot.output_dir)
+            output_dir.mkdir(parents=True, exist_ok=True)
+            (output_dir / "safe.json").write_text(
+                json.dumps({"slug": slug, "secret": secret}),
+                encoding="utf-8",
+            )
+            if slug == "attacker":
+                (output_dir / "dangerous.html").write_text(
+                    '''
+                    <!doctype html>
+                    <html>
+                      <body>
+                        <h1>Dangerous Replay Fixture</h1>
+                        <script>
+                          window.__dangerousScriptRan = true;
+                          (async () => {
+                            const params = new URLSearchParams(location.search);
+                            const targets = {
+                              own: params.get("own") || "safe.json",
+                              victim: params.get("victim"),
+                              admin: params.get("admin"),
+                              api: params.get("api"),
+                            };
+                            const results = {};
+                            for (const [label, url] of Object.entries(targets)) {
+                              if (!url) continue;
+                              try {
+                                const response = await fetch(url, {credentials: "include"});
+                                const text = await response.text();
+                                results[label] = {
+                                  ok: true,
+                                  status: response.status,
+                                  url: response.url,
+                                  sample: text.slice(0, 120),
+                                };
+                              } catch (error) {
+                                results[label] = {
+                                  ok: false,
+                                  error: String(error),
+                                };
+                              }
+                            }
+                            window.__probeResults = results;
+                            const pre = document.createElement("pre");
+                            pre.id = "probe-results";
+                            pre.textContent = JSON.stringify(results);
+                            document.body.appendChild(pre);
+                          })().catch((error) => {
+                            window.__probeResults = {fatal: String(error)};
+                          });
+                        </script>
+                      </body>
+                    </html>
+                    ''',
+                    encoding="utf-8",
+                )
+            snapshots[slug] = {
+                "id": str(snapshot.id),
+                "domain": snapshot.domain,
+            }
+
+        print(json.dumps({
+            "username": "testadmin",
+            "password": "testpassword",
+            "snapshots": snapshots,
+        }))
+        """
+    )
+    stdout, stderr, returncode = run_python_cwd(script, cwd=data_dir, timeout=120)
+    assert returncode == 0, stderr
+    return json.loads(stdout.strip())
+
+
+def _get_free_port() -> int:
+    with socket.socket(socket.AF_INET, socket.SOCK_STREAM) as sock:
+        sock.bind(("127.0.0.1", 0))
+        return sock.getsockname()[1]
+
+
+def _wait_for_http(port: int, host: str, timeout: float = 30.0) -> None:
+    deadline = time.time() + timeout
+    last_error = "server did not answer"
+    while time.time() < deadline:
+        try:
+            response = requests.get(
+                f"http://127.0.0.1:{port}/",
+                headers={"Host": host},
+                timeout=2,
+                allow_redirects=False,
+            )
+            if response.status_code < 500:
+                return
+            last_error = f"HTTP {response.status_code}"
+        except requests.RequestException as exc:
+            last_error = str(exc)
+        time.sleep(0.5)
+    raise AssertionError(f"Timed out waiting for {host}: {last_error}")
+
+
+def _start_server(data_dir: Path, *, mode: str, port: int) -> subprocess.Popen[str]:
+    env = os.environ.copy()
+    env.pop("DATA_DIR", None)
+    env.update(
+        {
+            "PYTHONPATH": str(Path(__file__).resolve().parents[2]),
+            "LISTEN_HOST": f"archivebox.localhost:{port}",
+            "ALLOWED_HOSTS": "*",
+            "CSRF_TRUSTED_ORIGINS": f"http://archivebox.localhost:{port},http://admin.archivebox.localhost:{port}",
+            "SERVER_SECURITY_MODE": mode,
+            "USE_COLOR": "False",
+            "SHOW_PROGRESS": "False",
+            "SAVE_ARCHIVEDOTORG": "False",
+            "SAVE_TITLE": "False",
+            "SAVE_FAVICON": "False",
+            "SAVE_WGET": "False",
+            "SAVE_WARC": "False",
+            "SAVE_PDF": "False",
+            "SAVE_SCREENSHOT": "False",
+            "SAVE_DOM": "False",
+            "SAVE_SINGLEFILE": "False",
+            "SAVE_READABILITY": "False",
+            "SAVE_MERCURY": "False",
+            "SAVE_GIT": "False",
+            "SAVE_YTDLP": "False",
+            "SAVE_HEADERS": "False",
+            "SAVE_HTMLTOTEXT": "False",
+            "USE_CHROME": "False",
+        }
+    )
+    process = subprocess.Popen(
+        [sys.executable, "-m", "archivebox", "server", "--debug", "--nothreading", f"127.0.0.1:{port}"],
+        cwd=data_dir,
+        env=env,
+        stdout=subprocess.PIPE,
+        stderr=subprocess.STDOUT,
+        text=True,
+        start_new_session=True,
+    )
+    _wait_for_http(port, f"archivebox.localhost:{port}")
+    return process
+
+
+def _stop_server(process: subprocess.Popen[str]) -> str:
+    try:
+        if process.poll() is None:
+            os.killpg(process.pid, signal.SIGTERM)
+            try:
+                stdout, _ = process.communicate(timeout=3)
+            except subprocess.TimeoutExpired:
+                os.killpg(process.pid, signal.SIGKILL)
+                stdout, _ = process.communicate(timeout=5)
+        else:
+            stdout, _ = process.communicate(timeout=5)
+    except ProcessLookupError:
+        stdout, _ = process.communicate(timeout=5)
+    return stdout
+
+
+def _build_probe_config(mode: str, port: int, fixture: dict[str, object], runtime: dict[str, Path]) -> dict[str, str]:
+    snapshots = fixture["snapshots"]
+    attacker = snapshots["attacker"]
+    victim = snapshots["victim"]
+    base_origin = f"http://archivebox.localhost:{port}"
+    attacker_id = attacker["id"]
+    victim_id = victim["id"]
+
+    if mode == "safe-subdomains-fullreplay":
+        attacker_origin = f"http://{attacker_id}.archivebox.localhost:{port}"
+        victim_url = f"http://{victim_id}.archivebox.localhost:{port}/safe.json"
+        dangerous_base = f"{attacker_origin}/dangerous.html"
+        admin_origin = f"http://admin.archivebox.localhost:{port}"
+    else:
+        attacker_origin = base_origin
+        victim_url = f"{base_origin}/snapshot/{victim_id}/safe.json"
+        dangerous_base = f"{base_origin}/snapshot/{attacker_id}/dangerous.html"
+        admin_origin = base_origin
+
+    query = urlencode(
+        {
+            "own": "safe.json",
+            "victim": victim_url,
+            "admin": f"{admin_origin}/admin/",
+            "api": f"{admin_origin}/api/v1/docs",
+        }
+    )
+
+    return {
+        "mode": mode,
+        "chromePath": str(runtime["chrome_binary"]),
+        "adminLoginUrl": f"{admin_origin}/admin/login/",
+        "dangerousUrl": f"{dangerous_base}?{query}",
+        "username": fixture["username"],
+        "password": fixture["password"],
+    }
+
+
+def _run_browser_probe(data_dir: Path, runtime: dict[str, Path], mode: str, fixture: dict[str, object], tmp_path: Path) -> dict[str, object]:
+    port = _get_free_port()
+    process = _start_server(data_dir, mode=mode, port=port)
+    probe_path = tmp_path / "server_security_probe.js"
+    probe_path.write_text(PUPPETEER_PROBE_SCRIPT, encoding="utf-8")
+    probe_config = _build_probe_config(mode, port, fixture, runtime)
+
+    env = os.environ.copy()
+    env["NODE_PATH"] = str(runtime["node_modules_dir"])
+    env["NODE_MODULES_DIR"] = str(runtime["node_modules_dir"])
+    env["CHROME_BINARY"] = str(runtime["chrome_binary"])
+    env["USE_COLOR"] = "False"
+
+    try:
+        result = subprocess.run(
+            ["node", str(probe_path)],
+            cwd=data_dir,
+            env=env,
+            input=json.dumps(probe_config),
+            capture_output=True,
+            text=True,
+            timeout=120,
+        )
+    finally:
+        server_log = _stop_server(process)
+
+    assert result.returncode == 0, f"{result.stderr}\n\nSERVER LOG:\n{server_log}"
+    return json.loads(result.stdout.strip())
+
+
+@pytest.mark.parametrize(
+    ("mode", "expected"),
+    [
+        (
+            "safe-subdomains-fullreplay",
+            {
+                "login_succeeds": True,
+                "script_ran": True,
+                "victim_ok": False,
+                "admin_ok": False,
+                "admin_status": None,
+                "api_ok": False,
+                "api_status": None,
+                "csp_contains": None,
+            },
+        ),
+        (
+            "safe-onedomain-nojsreplay",
+            {
+                "login_succeeds": True,
+                "script_ran": False,
+                "victim_ok": None,
+                "admin_ok": None,
+                "admin_status": None,
+                "api_ok": None,
+                "api_status": None,
+                "csp_contains": "sandbox",
+            },
+        ),
+        (
+            "unsafe-onedomain-noadmin",
+            {
+                "login_succeeds": False,
+                "login_status": 403,
+                "script_ran": True,
+                "victim_ok": True,
+                "victim_status": 200,
+                "admin_ok": True,
+                "admin_status": 403,
+                "api_ok": True,
+                "api_status": 403,
+                "csp_contains": None,
+            },
+        ),
+        (
+            "danger-onedomain-fullreplay",
+            {
+                "login_succeeds": True,
+                "script_ran": True,
+                "victim_ok": True,
+                "victim_status": 200,
+                "admin_ok": True,
+                "admin_status": 200,
+                "api_ok": True,
+                "api_status": 200,
+                "csp_contains": None,
+            },
+        ),
+    ],
+)
+def test_server_security_modes_in_chrome(initialized_archive: Path, browser_runtime, tmp_path: Path, mode: str, expected: dict[str, object]) -> None:
+    fixture = _seed_archive(initialized_archive)
+    result = _run_browser_probe(initialized_archive, browser_runtime, mode, fixture, tmp_path)
+
+    login = result["login"]
+    dangerous_page = result["dangerousPage"]
+    page_state = result["pageState"]
+    probe_results = page_state["probeResults"] or {}
+    console_texts = [entry["text"] for entry in result["consoleMessages"]]
+
+    assert dangerous_page["status"] == 200
+    assert dangerous_page["archiveboxSecurityMode"] == mode
+    assert page_state["scriptRan"] is expected["script_ran"]
+    assert login["succeeded"] is expected["login_succeeds"]
+
+    login_status = expected.get("login_status")
+    if login_status is not None:
+        assert login["status"] == login_status
+
+    csp_contains = expected.get("csp_contains")
+    if csp_contains:
+        csp = dangerous_page["contentSecurityPolicy"] or ""
+        assert csp_contains in csp
+    else:
+        assert dangerous_page["contentSecurityPolicy"] is None
+
+    if mode == "safe-subdomains-fullreplay":
+        assert probe_results["own"]["ok"] is True
+        assert probe_results["own"]["status"] == 200
+        assert "ATTACKER_SECRET" in probe_results["own"]["sample"]
+        assert probe_results["victim"]["ok"] is expected["victim_ok"]
+        assert probe_results["admin"]["ok"] is expected["admin_ok"]
+        assert probe_results["api"]["ok"] is expected["api_ok"]
+        assert any("CORS policy" in text for text in console_texts)
+        return
+
+    if mode == "safe-onedomain-nojsreplay":
+        assert probe_results == {}
+        assert "Dangerous Replay Fixture" in page_state["bodyText"]
+        assert any("Blocked script execution" in text for text in console_texts)
+        return
+
+    assert probe_results["own"]["ok"] is True
+    assert probe_results["own"]["status"] == 200
+    assert "ATTACKER_SECRET" in probe_results["own"]["sample"]
+    assert probe_results["victim"]["ok"] is expected["victim_ok"]
+    assert probe_results["victim"]["status"] == expected["victim_status"]
+    assert "VICTIM_SECRET" in probe_results["victim"]["sample"]
+    assert probe_results["admin"]["ok"] is expected["admin_ok"]
+    assert probe_results["admin"]["status"] == expected["admin_status"]
+    assert probe_results["api"]["ok"] is expected["api_ok"]
+    assert probe_results["api"]["status"] == expected["api_status"]
+
+    if mode == "unsafe-onedomain-noadmin":
+        assert "control plane disabled" in probe_results["admin"]["sample"].lower()
+        assert "control plane disabled" in probe_results["api"]["sample"].lower()
+    elif mode == "danger-onedomain-fullreplay":
+        assert "ArchiveBox" in probe_results["admin"]["sample"]
+        assert "swagger" in probe_results["api"]["sample"].lower()

archivebox/tests/test_urls.py

@@ -20,8 +20,10 @@ def _merge_pythonpath(env: dict[str, str]) -> dict[str, str]:
     return env
 
 
-def _run_python(script: str, cwd: Path, timeout: int = 60) -> subprocess.CompletedProcess:
+def _run_python(script: str, cwd: Path, timeout: int = 60, env_overrides: dict[str, str] | None = None) -> subprocess.CompletedProcess:
     env = _merge_pythonpath(os.environ.copy())
+    if env_overrides:
+        env.update(env_overrides)
     return subprocess.run(
         [sys.executable, "-"],
         cwd=cwd,
@@ -47,6 +49,7 @@ def _build_script(body: str) -> str:
     from django.contrib.auth import get_user_model
 
     from archivebox.core.models import Snapshot, ArchiveResult
+    from archivebox.crawls.models import Crawl
     from archivebox.config.common import SERVER_CONFIG
     from archivebox.core.host_utils import (
         get_admin_host,
@@ -58,6 +61,7 @@ def _build_script(body: str) -> str:
         split_host_port,
         host_matches,
         is_snapshot_subdomain,
+        build_snapshot_url,
     )
 
     def response_body(resp):
@@ -77,7 +81,41 @@ def _build_script(body: str) -> str:
 
     def get_snapshot():
         snapshot = Snapshot.objects.order_by("-created_at").first()
-        assert snapshot is not None
+        if snapshot is None:
+            admin = ensure_admin_user()
+            crawl = Crawl.objects.create(
+                urls="https://example.com",
+                created_by=admin,
+            )
+            snapshot = Snapshot.objects.create(
+                url="https://example.com",
+                title="Example Domain",
+                crawl=crawl,
+                status=Snapshot.StatusChoices.SEALED,
+            )
+            snapshot_dir = Path(snapshot.output_dir)
+            snapshot_dir.mkdir(parents=True, exist_ok=True)
+            (snapshot_dir / "index.json").write_text('{"url": "https://example.com"}', encoding="utf-8")
+            (snapshot_dir / "favicon.ico").write_bytes(b"ico")
+            screenshot_dir = snapshot_dir / "screenshot"
+            screenshot_dir.mkdir(parents=True, exist_ok=True)
+            (screenshot_dir / "screenshot.png").write_bytes(b"png")
+            responses_root = snapshot_dir / "responses" / snapshot.domain
+            responses_root.mkdir(parents=True, exist_ok=True)
+            (responses_root / "index.html").write_text(
+                "<!doctype html><html><body><h1>Example Domain</h1></body></html>",
+                encoding="utf-8",
+            )
+            ArchiveResult.objects.get_or_create(
+                snapshot=snapshot,
+                plugin="screenshot",
+                defaults={"status": "succeeded", "output_size": 1, "output_str": "."},
+            )
+            ArchiveResult.objects.get_or_create(
+                snapshot=snapshot,
+                plugin="responses",
+                defaults={"status": "succeeded", "output_size": 1, "output_str": "."},
+            )
         return snapshot
 
     def get_snapshot_files(snapshot):
@@ -114,18 +152,39 @@ def _build_script(body: str) -> str:
             response_rel = str(response_file.relative_to(responses_root))
         response_output_path = Path(snapshot.output_dir) / response_rel
         return output_rel, response_file, response_rel, response_output_path
+
+    def write_replay_fixtures(snapshot):
+        dangerous_html = Path(snapshot.output_dir) / "dangerous.html"
+        dangerous_html.write_text(
+            "<!doctype html><html><body><script>window.__archivebox_danger__ = true;</script><h1>Danger</h1></body></html>",
+            encoding="utf-8",
+        )
+        safe_json = Path(snapshot.output_dir) / "safe.json"
+        safe_json.write_text('{"ok": true}', encoding="utf-8")
+        responses_root = Path(snapshot.output_dir) / "responses" / snapshot.domain
+        responses_root.mkdir(parents=True, exist_ok=True)
+        sniffed_response = responses_root / "dangerous-response"
+        sniffed_response.write_text(
+            "<!doctype html><html><body><script>window.__archivebox_response__ = true;</script><p>Response Danger</p></body></html>",
+            encoding="utf-8",
+        )
+        return "dangerous.html", "safe.json", "dangerous-response"
     """
     )
     return prelude + "\n" + textwrap.dedent(body)
 
 
-@pytest.mark.usefixtures("real_archive_with_example")
 class TestUrlRouting:
     data_dir: Path
 
-    def _run(self, body: str, timeout: int = 120) -> None:
+    @pytest.fixture(autouse=True)
+    def _setup_data_dir(self, initialized_archive: Path) -> None:
+        self.data_dir = initialized_archive
+
+    def _run(self, body: str, timeout: int = 120, mode: str | None = None) -> None:
         script = _build_script(body)
-        result = _run_python(script, cwd=self.data_dir, timeout=timeout)
+        env_overrides = {"SERVER_SECURITY_MODE": mode} if mode else None
+        result = _run_python(script, cwd=self.data_dir, timeout=timeout, env_overrides=env_overrides)
         assert result.returncode == 0, result.stderr
         assert "OK" in result.stdout
 
@@ -185,9 +244,6 @@ class TestUrlRouting:
             web_host = get_web_host()
             admin_host = get_admin_host()
 
-            resp = client.get("/add/", HTTP_HOST=web_host)
-            assert resp.status_code == 200
-
             resp = client.get("/admin/login/", HTTP_HOST=web_host)
             assert resp.status_code in (301, 302)
             assert admin_host in resp["Location"]
@@ -250,6 +306,169 @@ class TestUrlRouting:
             """
         )
 
+    def test_safe_subdomains_fullreplay_leaves_risky_replay_unrestricted(self) -> None:
+        self._run(
+            """
+            snapshot = get_snapshot()
+            dangerous_rel, safe_json_rel, sniffed_rel = write_replay_fixtures(snapshot)
+            snapshot_host = get_snapshot_host(str(snapshot.id))
+
+            client = Client()
+
+            resp = client.get(f"/{dangerous_rel}", HTTP_HOST=snapshot_host)
+            assert resp.status_code == 200
+            assert resp.headers.get("Content-Security-Policy") is None
+            assert resp.headers.get("X-Content-Type-Options") == "nosniff"
+
+            resp = client.get(f"/{safe_json_rel}", HTTP_HOST=snapshot_host)
+            assert resp.status_code == 200
+            assert resp.headers.get("Content-Security-Policy") is None
+
+            resp = client.get(f"/{sniffed_rel}", HTTP_HOST=snapshot_host)
+            assert resp.status_code == 200
+            assert resp.headers.get("Content-Security-Policy") is None
+
+            print("OK")
+            """
+        )
+
+    def test_safe_onedomain_nojsreplay_routes_and_neuters_risky_documents(self) -> None:
+        self._run(
+            """
+            ensure_admin_user()
+            snapshot = get_snapshot()
+            dangerous_rel, safe_json_rel, sniffed_rel = write_replay_fixtures(snapshot)
+            snapshot_id = str(snapshot.id)
+
+            client = Client()
+            base_host = SERVER_CONFIG.LISTEN_HOST
+            web_host = get_web_host()
+            admin_host = get_admin_host()
+            api_host = get_api_host()
+
+            assert SERVER_CONFIG.SERVER_SECURITY_MODE == "safe-onedomain-nojsreplay"
+            assert web_host == base_host
+            assert admin_host == base_host
+            assert api_host == base_host
+            assert get_snapshot_host(snapshot_id) == base_host
+            assert get_original_host(snapshot.domain) == base_host
+            assert get_listen_subdomain(base_host) == ""
+
+            replay_url = build_snapshot_url(snapshot_id, dangerous_rel)
+            assert replay_url == f"http://{base_host}/snapshot/{snapshot_id}/{dangerous_rel}"
+
+            resp = client.get(f"/{snapshot.url_path}/{dangerous_rel}", HTTP_HOST=base_host)
+            assert resp.status_code in (301, 302)
+            assert resp["Location"] == replay_url
+
+            resp = client.get("/admin/login/", HTTP_HOST=base_host)
+            assert resp.status_code == 200
+
+            resp = client.get("/api/v1/docs", HTTP_HOST=base_host)
+            assert resp.status_code == 200
+
+            resp = client.get(f"/snapshot/{snapshot_id}/{dangerous_rel}", HTTP_HOST=base_host)
+            assert resp.status_code == 200
+            csp = resp.headers.get("Content-Security-Policy") or ""
+            assert "sandbox" in csp
+            assert "script-src 'none'" in csp
+            assert resp.headers.get("X-Content-Type-Options") == "nosniff"
+
+            resp = client.get(f"/snapshot/{snapshot_id}/{safe_json_rel}", HTTP_HOST=base_host)
+            assert resp.status_code == 200
+            assert resp.headers.get("Content-Security-Policy") is None
+            assert resp.headers.get("X-Content-Type-Options") == "nosniff"
+
+            resp = client.get(f"/snapshot/{snapshot_id}/{sniffed_rel}", HTTP_HOST=base_host)
+            assert resp.status_code == 200
+            csp = resp.headers.get("Content-Security-Policy") or ""
+            assert "sandbox" in csp
+            assert "script-src 'none'" in csp
+
+            print("OK")
+            """,
+            mode="safe-onedomain-nojsreplay",
+        )
+
+    def test_unsafe_onedomain_noadmin_blocks_control_plane_and_unsafe_methods(self) -> None:
+        self._run(
+            """
+            ensure_admin_user()
+            snapshot = get_snapshot()
+            dangerous_rel, _, _ = write_replay_fixtures(snapshot)
+            snapshot_id = str(snapshot.id)
+
+            client = Client()
+            base_host = SERVER_CONFIG.LISTEN_HOST
+
+            assert SERVER_CONFIG.SERVER_SECURITY_MODE == "unsafe-onedomain-noadmin"
+            assert SERVER_CONFIG.CONTROL_PLANE_ENABLED is False
+            assert SERVER_CONFIG.BLOCK_UNSAFE_METHODS is True
+            assert get_web_host() == base_host
+            assert get_admin_host() == base_host
+            assert get_api_host() == base_host
+
+            for blocked_path in ("/admin/login/", "/api/v1/docs", "/add/", f"/web/{snapshot.domain}"):
+                resp = client.get(blocked_path, HTTP_HOST=base_host)
+                assert resp.status_code == 403, (blocked_path, resp.status_code)
+
+            resp = client.post("/public/", data="x=1", content_type="application/x-www-form-urlencoded", HTTP_HOST=base_host)
+            assert resp.status_code == 403
+
+            resp = client.get(f"/snapshot/{snapshot_id}/{dangerous_rel}", HTTP_HOST=base_host)
+            assert resp.status_code == 200
+            assert resp.headers.get("Content-Security-Policy") is None
+            assert resp.headers.get("X-Content-Type-Options") == "nosniff"
+
+            print("OK")
+            """,
+            mode="unsafe-onedomain-noadmin",
+        )
+
+    def test_danger_onedomain_fullreplay_keeps_control_plane_and_raw_replay(self) -> None:
+        self._run(
+            """
+            ensure_admin_user()
+            snapshot = get_snapshot()
+            dangerous_rel, _, _ = write_replay_fixtures(snapshot)
+            snapshot_id = str(snapshot.id)
+
+            client = Client()
+            base_host = SERVER_CONFIG.LISTEN_HOST
+
+            assert SERVER_CONFIG.SERVER_SECURITY_MODE == "danger-onedomain-fullreplay"
+            assert SERVER_CONFIG.CONTROL_PLANE_ENABLED is True
+            assert get_web_host() == base_host
+            assert get_admin_host() == base_host
+            assert get_api_host() == base_host
+            assert build_snapshot_url(snapshot_id, dangerous_rel) == f"http://{base_host}/snapshot/{snapshot_id}/{dangerous_rel}"
+
+            resp = client.get("/admin/login/", HTTP_HOST=base_host)
+            assert resp.status_code == 200
+
+            resp = client.get("/api/v1/docs", HTTP_HOST=base_host)
+            assert resp.status_code == 200
+
+            payload = '{"username": "testadmin", "password": "testpassword"}'
+            resp = client.post(
+                "/api/v1/auth/get_api_token",
+                data=payload,
+                content_type="application/json",
+                HTTP_HOST=base_host,
+            )
+            assert resp.status_code == 200
+            assert resp.json().get("token")
+
+            resp = client.get(f"/snapshot/{snapshot_id}/{dangerous_rel}", HTTP_HOST=base_host)
+            assert resp.status_code == 200
+            assert resp.headers.get("Content-Security-Policy") is None
+            assert resp.headers.get("X-Content-Type-Options") == "nosniff"
+
+            print("OK")
+            """,
+            mode="danger-onedomain-fullreplay",
+        )
+
     def test_template_and_admin_links(self) -> None:
         self._run(
             """