A per-request policy build (forced by the nonce) means a stray semicolon in
a CSP_EXTRA_* value — a plausible operator paste like
"https://youtube.com; https://vimeo.com" — lands inside a single source
token and makes Rails raise InvalidDirectiveError on every request, a
site-wide 500 even in report-only mode. Split on ';' alongside comma and
whitespace so such a value tokenizes into valid sources. Rails still
validates each token, so no injection is introduced (a token with an
embedded space still fails validation).
Campfire is self-hosted per customer: an admin may embed or connect to
external hosts (video/embed providers, image CDNs, analytics, form or
webhook endpoints) that vary per install and are unknown at build time.
Hardcoding these directives at :self would break those integrations once
the policy is enforced.
Read per-install extras from ENV (CSP_EXTRA_SCRIPT_SRC, _STYLE_SRC,
_IMG_SRC, _CONNECT_SRC, _FRAME_SRC, _FORM_ACTION), each a comma- or
whitespace-separated host list, appended to the :self baseline. Unset by
default, so the policy stays at :self only unless an admin opts in.
Layer a Turbo-safe nonce and a full directive set onto the Report-Only
baseline, still without enforcing anything.
The nonce is the HMAC-SHA256 of a stable per-visitor cookie, keyed by
secret_key_base. It stays constant across a session's requests so Turbo
snapshot restores don't replay a stale nonce and trip the policy, while
staying unpredictable to a client that can set the cookie but not the
secret. importmap-rails auto-nonces the importmap JSON and shim, the only
inline scripts the app renders (the message _template is a non-executed
text/template data block).
script-src gains the nonce; style-src keeps unsafe_inline for now (inline
style= attributes across the composer and ActionText content aren't
nonceable yet); img/connect/frame-src start at :self as tuning starting
points for the report-only window. report_only stays true; enforcement is
a later flip.
UnreadRoomsChannel streamed from a hardcoded global name, and every message
published its room id to it. Any authenticated user, including one with no
memberships at all, could subscribe and watch which rooms were active and exactly
when, across every closed room and direct conversation on the account. The browser
filters ids it doesn't recognise, but that happens after delivery.
Its sibling ReadRoomsChannel is already scoped per user; this mirrors it, and
message creation fans the notice out to the room's members instead of broadcasting
it to everyone. No message content was exposed either way, only the timing.
Message content is delivered over turbo streams, which ran on the stock
Turbo::StreamsChannel. That channel verifies the signature on the stream name and
nothing else. The name carries no expiry and no binding to a user, so one read off
the page while a member kept working after the membership was revoked.
Revocation made this worse rather than better. Membership#after_destroy_commit
disconnects the user with reconnect: true, and the client replays its subscriptions
on the new socket: RoomChannel re-checks membership and rejects, while the turbo
subscription re-verified only the signature and was accepted.
RoomMessagesChannel re-checks membership on every subscribe, deriving the room from
the verified stream name so there is no parameter to point elsewhere. Since the
subscriber names the channel it wants, the stock channel would otherwise be a way
around that check, so it now turns these stream names away and this is the only door.
Rooms::DirectsController relaxes ensure_can_administer to true, because every
participant in a direct room may administer it. set_room was inherited unscoped,
though, so that relaxation applied to any room the caller was merely a member of:
DELETE /rooms/directs/<id> destroyed open and closed rooms and all their messages.
The same unscoped lookup let a direct room be loaded by the opens and closeds
controllers, where force_room_type promoted it. Promoting a DM to open grants every
user on the account membership and republishes the whole conversation, including the
other participant's messages; converting it to closed lets the initiator revise who
is in it and lock the other participant out.
Each controller now narrows room_scope to the types it may act on. Opens and closeds
keep reach into each other, since converting between them is a feature. Neither can
reach a direct room, and directs can only reach directs.
Room also refuses to change type away from Rooms::Direct, so the invariant holds for
any future caller of becomes! rather than only these two controllers.
and add test coverage for (un)supported file types.
The avatar and logo variants move into the models and return nil for content
types that are no longer variable, so the controllers fall back to the initials
avatar and stock logo icon instead of raising `ActiveStorage::InvariableError`.
The guard uses IPAddr but relied on something else loading it first;
require it explicitly. And the rebinding tests lost their port assertion
when the matchers were loosened for newer Net::HTTP keyword args, so
check the port alongside the IP again.
Matches the fizzy guard: the local-use NAT64 prefix (64:ff9b:1::/48,
RFC 8215) embeds an IPv4 target in its low 32 bits just like the
well-known prefix, so run it through the same embedded-IPv4 recheck.
Local-use NAT64 to a public address now resolves (keeping unfurls
working for self-hosters on such networks) while local-use NAT64 to an
internal address stays blocked.
The local-use NAT64 prefix (64:ff9b:1::/48, RFC 8215) embeds an IPv4
target like the well-known prefix does, but at a deployment-chosen
position we can't extract, so block the whole range outright.
Also block the IPv6 benchmarking range (2001:2::/48, RFC 5180) to match
the IPv4 benchmarking block on 198.18.0.0/15.
Net::HTTP now passes an open_timeout: option to TCPSocket.open, so the mock
that matched exact positional arguments no longer matches. Match on the
host instead.
The guard blocked the usual private, loopback, and link-local ranges (and
the IPv4-mapped/-compatible IPv6 forms), but let through NAT64, 6to4, and
Teredo addresses, which can point at an internal IPv4, and CGNAT.
Now it pulls the IPv4 out of a NAT64 address and checks that (so NAT64 to a
public site still works), blocks 6to4 and Teredo outright, and adds the
missing IPv4 and IPv6 ranges.
Adds ipv4_mapped? and ipv4_compat? checks to PrivateNetworkGuard.private_ip?
to block SSRF bypass attempts using IPv6 address formats like:
- ::ffff:169.254.169.254 (IPv4-mapped)
- ::169.254.169.254 (IPv4-compatible)
These formats could previously bypass the link_local? check since Ruby
treats them as IPv6 addresses, not IPv4.
Ref: HackerOne #3481701
This adds the ability to ban a user by their IP address.
When an admin is viewing a user profile, a new "Ban user" button is
present. Clicking on that will:
- Create a ban on the IP addresses that were tracked for that user's
sessions
- Remove all the messages authored by that user
- Log the user out immediately
In addition, that user will no longer be shown in most user lists in the
app. They are still shown to admins, in account settings. Viewing their
profile from there will now show a "Remove ban" button which can be used
to restore their access (it doesn't restore their messages though --
those are already gone -- it just removes the blocks so they can log in
again).
If bot_key has no right-hand side (ex: 1-), bot_token will be nil, and the query will match a User record if bot_id matches a valid ID.
Fix it relying on `active_bots` instead.
* Bump Ruby to 3.4.5
* Update dependencies
* Adjust for Rails 8 and Ruby 3.5 API changes
* Mark params strings as mutable in prepapration for frozen strings in Ruby 3.5
* Update test for HTML5 sanitizer
With Rails 7.1 the HTML5 sanitizer became the default, this breakts this test because the old sanitizer used to delete unpermitted nodes, while the new one returns their content
The final string is safe, but different then it used to be in Rails 7.0
* Remove direct Turbo tesh helpers require & parallelize tests
* Fix Zeitwerk issues with rails extensions
* Update Resque setup for Redis 5+
* Remove unused views
* Remove GID v1 handler
Jeremy Daer
Mike Dalessio
Donal McBreen
Rosa Gutierrez
Ashwin M
Michael Halliday
Stanko K.R.
David Heinemeier Hansson
Kevin McConnell
Jacopo