Vulnerabilities
CVE-2022-23549
MEDIUMDiscourse is an option source discussion platform. Prior to version 2.8.14 on the `stable` branch and version 2.9.0.beta16 on the `beta` and `tests-passed` branches, users can create posts with raw body longer than the `max_length` site setting by including html comments that are not counted toward the character limit. This issue is patched in versions 2.8.14 and 2.9.0.beta16. There are no known workarounds.
CVE-2022-23548
MEDIUMDiscourse is an option source discussion platform. Prior to version 2.8.14 on the `stable` branch and version 2.9.0.beta16 on the `beta` and `tests-passed` branches, parsing posts can be susceptible to regular expression denial of service (ReDoS) attacks. This issue is patched in versions 2.8.14 and 2.9.0.beta16. There are no known workarounds.
CVE-2022-46180
MEDIUMDiscourse Mermaid (discourse-mermaid-theme-component) allows users of Discourse, open-source forum software, to create graphs using the Mermaid syntax. Users of discourse-mermaid-theme-component version 1.0.0 who can create posts are able to inject arbitrary HTML on that post. The issue has been fixed on the `main` branch of the GitHub repository, with 1.1.0 named as a patched version. Admins can update the theme component through the admin UI. As a workaround, admins can temporarily disable discourse-mermaid-theme-component.
CVE-2022-46159
MEDIUMDiscourse is an open-source discussion platform. In version 2.8.13 and prior on the `stable` branch and version 2.9.0.beta14 and prior on the `beta` and `tests-passed` branches, any authenticated user can create an unlisted topic. These topics, which are not readily available to other users, can take up unnecessary site resources. A patch for this issue is available in the `main` branch of Discourse. There are no known workarounds available.
CVE-2022-46162
HIGHdiscourse-bbcode is the official BBCode plugin for Discourse. Prior to commit 91478f5, CSS injection can occur when rendering content generated with the discourse-bccode plugin. This vulnerability only affects sites which have the discourse-bbcode plugin installed and enabled. This issue is patched in commit 91478f5. As a workaround, ensure that the Content Security Policy is enabled and monitor any posts that contain bbcode.
CVE-2022-46148
HIGHDiscourse is an open-source messaging platform. In versions 2.8.10 and prior on the `stable` branch and versions 2.9.0.beta11 and prior on the `beta` and `tests-passed` branches, users composing malicious messages and navigating to drafts page could self-XSS. This vulnerability can lead to a full XSS on sites which have modified or disabled Discourse’s default Content Security Policy. This issue is patched in the latest stable, beta and tests-passed versions of Discourse.
CVE-2022-46150
MEDIUMDiscourse is an open-source discussion platform. Prior to version 2.8.13 of the `stable` branch and version 2.9.0.beta14 of the `beta` and `tests-passed` branches, unauthorized users may learn of the existence of hidden tags and that they have been applied to topics that they have access to. This issue is patched in version 2.8.13 of the `stable` branch and version 2.9.0.beta14 of the `beta` and `tests-passed` branches. As a workaround, use the `disable_email` site setting to disable all emails to non-staff users.
CVE-2022-41944
LOWDiscourse is an open-source discussion platform. In stable versions prior to 2.8.12 and beta or tests-passed versions prior to 2.9.0.beta.13, under certain conditions, a user can see notifications for topics they no longer have access to. If there is sensitive information in the topic title, it will therefore have been exposed. This issue is patched in stable version 2.8.12, beta version 2.9.0.beta13, and tests-passed version 2.9.0.beta13. There are no workarounds available.
CVE-2022-41921
LOWDiscourse is an open-source discussion platform. Prior to version 2.9.0.beta13, users can post chat messages of an unlimited length, which can cause a denial of service for other users when posting huge amounts of text. Users should upgrade to version 2.9.0.beta13, where a limit has been introduced. No known workarounds are available.
CVE-2022-41913
MEDIUMDiscourse-calendar is a plugin for the Discourse messaging platform which adds the ability to create a dynamic calendar in the first post of a topic. Members of private groups or public groups with private members can be listed by users, who can create and edit post events. This vulnerability only affects sites which have discourse post events enabled. This issue has been patched in commit `ca5ae3e7e` which will be included in future releases. Users unable to upgrade should disable the `discourse_post_event_enabled` setting to fully mitigate the issue. Also, it's possible to prevent regular users from using this vulnerability by removing all groups from the `discourse_post_event_allowed_on_groups` but note that moderators will still be able to use it.
CVE-2022-39385
MEDIUMDiscourse is the an open source discussion platform. In some rare cases users redeeming an invitation can be added as a participant to several private message topics that they should not be added to. They are not notified of this, it happens transparently in the background. This issue has been resolved in commit `a414520742` and will be included in future releases. Users are advised to upgrade. Users are also advised to set `SiteSetting.max_invites_per_day` to 0 until the patch is installed.
CVE-2022-39241
HIGHDiscourse is a platform for community discussion. A malicious admin could use this vulnerability to perform port enumeration on the local host or other hosts on the internal network, as well as against hosts on the Internet. Latest `stable`, `beta`, and `test-passed` versions are now patched. As a workaround, self-hosters can use `DISCOURSE_BLOCKED_IP_BLOCKS` env var (which overrides `blocked_ip_blocks` setting) to stop webhooks from accessing private IPs.
CVE-2022-39356
HIGHDiscourse is a platform for community discussion. Users who receive an invitation link that is not scoped to a single email address can enter any non-admin user's email and gain access to their account when accepting the invitation. All users should upgrade to the latest version. A workaround is temporarily disabling invitations with `SiteSetting.max_invites_per_day = 0` or scope them to individual email addresses.
CVE-2022-39378
MEDIUMDiscourse is a platform for community discussion. Under certain conditions, a user badge may have been awarded based on a user's activity in a topic with restricted access. Before this vulnerability was disclosed, the topic title of the topic associated with the user badge may be viewed by any user. If there are sensitive information in the topic title, it will therefore have been exposed. This issue is patched in the latest stable, beta and tests-passed versions of Discourse. There are currently no known workarounds available.
CVE-2022-39355
CRITICALDiscourse Patreon enables syncronization between Discourse Groups and Patreon rewards. On sites with Patreon login enabled, an improper authentication vulnerability could be used to take control of a victim's forum account. This vulnerability is patched in commit number 846d012151514b35ce42a1636c7d70f6dcee879e of the discourse-patreon plugin. Out of an abundance of caution, any Discourse accounts which have logged in with an unverified-email Patreon account will be logged out and asked to verify their email address on their next login. As a workaround, disable the patreon integration and log out all users with associated Patreon accounts.
CVE-2022-39279
MEDIUMdiscourse-chat is a plugin for the Discourse message board which adds chat functionality. In versions prior to 0.9 some places render a chat channel's name and description in an unsafe way, allowing staff members to cause an cross site scripting (XSS) attack by inserting unsafe HTML into them. Version 0.9 has addressed this issue. Users are advised to upgrade. There are no known workarounds for this issue.
CVE-2022-39270
MEDIUMDiscoTOC is a Discourse theme component that generates a table of contents for topics. Users that can create topics in TOC-enabled categories (and have sufficient trust level - configured in component's settings) are able to inject arbitrary HTML on that topic's page. The issue has been fixed on the `main` branch. Admins can update the theme component through the admin UI (Customize -> Themes -> Components -> DiscoTOC -> Check for Updates). Alternatively, admins can temporarily disable the DiscoTOC theme component.
CVE-2022-39232
MEDIUMDiscourse is an open source discussion platform. Starting with version 2.9.0.beta5 and prior to version 2.9.0.beta10, an incomplete quote can generate a JavaScript error which will crash the current page in the browser in some cases. Version 2.9.0.beta10 added a fix and tests to ensure incomplete quotes won't break the app. As a workaround, the quote can be fixed via the rails console.
CVE-2022-39226
MEDIUMDiscourse is an open source discussion platform. In versions prior to 2.8.9 on the `stable` branch and prior to 2.9.0.beta10 on the `beta` and `tests-passed` branches, a malicious actor can add large payloads of text into the Location and Website fields of a user profile, which causes issues for other users when loading that profile. A fix to limit the length of user input for these fields is included in version 2.8.9 on the `stable` branch and version 2.9.0.beta10 on the `beta` and `tests-passed` branches. There are no known workarounds.
CVE-2022-36068
HIGHDiscourse is an open source discussion platform. In versions prior to 2.8.9 on the `stable` branch and prior to 2.9.0.beta10 on the `beta` and `tests-passed` branches, a moderator can create new and edit existing themes by using the API when they should not be able to do so. The problem is patched in version 2.8.9 on the `stable` branch and version 2.9.0.beta10 on the `beta` and `tests-passed` branches. There are no known workarounds.
CVE-2022-36066
CRITICALDiscourse is an open source discussion platform. In versions prior to 2.8.9 on the `stable` branch and prior to 2.9.0.beta10 on the `beta` and `tests-passed` branches, admins can upload a maliciously crafted Zip or Gzip Tar archive to write files at arbitrary locations and trigger remote code execution. The problem is patched in version 2.8.9 on the `stable` branch and version 2.9.0.beta10 on the `beta` and `tests-passed` branches. There are no known workarounds.
CVE-2022-36057
MEDIUMDiscourse-Chat is an asynchronous messaging plugin for the Discourse open-source discussion platform. Users of Discourse Chat can be affected by admin users inserting HTML into chat titles and descriptions, causing a Cross-Site Scripting (XSS) attack. Version 0.9 contains a patch for this issue.
CVE-2022-31184
MEDIUMDiscourse is the an open source discussion platform. In affected versions an email activation route can be abused to send mass spam emails. A fix has been included in the latest stable, beta and tests-passed versions of Discourse which rate limits emails. Users are advised to upgrade. Users unable to upgrade should manually rate limit email.
CVE-2022-31182
MEDIUMDiscourse is the an open source discussion platform. In affected versions a maliciously crafted request for static assets could cause error responses to be cached by Discourse's default NGINX proxy configuration. A corrected NGINX configuration is included in the latest stable, beta and tests-passed versions of Discourse. Users are advised to upgrade. There are no known workarounds for this vulnerability.
CVE-2022-31096
MEDIUMDiscourse is an open source discussion platform. Under certain conditions, a logged in user can redeem an invite with an email that either doesn't match the invite's email or does not adhere to the email domain restriction of an invite link. The impact of this flaw is aggravated when the invite has been configured to add the user that accepts the invite into restricted groups. Once a user has been incorrectly added to a restricted group, the user may then be able to view content which that are restricted to the respective group. Users are advised to upgrade to the current stable releases. There are no known workarounds to this issue.
CVE-2022-31095
MEDIUMdiscourse-chat is a chat plugin for the Discourse application. Versions prior to 0.4 are vulnerable to an exposure of sensitive information, where an attacker who knows the message ID for a channel they do not have access to can view that message using the chat message lookup endpoint, primarily affecting direct message channels. There are no known workarounds for this issue, and users are advised to update the plugin.
CVE-2022-31060
MEDIUMDiscourse is an open-source discussion platform. Prior to version 2.8.4 in the `stable` branch and version `2.9.0.beta5` in the `beta` and `tests-passed` branches, banner topic data is exposed on login-required sites. This issue is patched in version 2.8.4 in the `stable` branch and version `2.9.0.beta5` in the `beta` and `tests-passed` branches of Discourse. As a workaround, one may disable banners.
CVE-2022-31059
MEDIUMDiscourse Calendar is a calendar plugin for Discourse, an open-source messaging app. Prior to version 1.0.1, parsing and rendering of Event names can be susceptible to cross-site scripting (XSS) attacks. This vulnerability only affects sites which have modified or disabled Discourse’s default Content Security Policy. This issue is patched in version 1.0.1 of the Discourse Calendar plugin. As a workaround, ensure that the Content Security Policy is enabled, and has not been modified in a way which would make it more vulnerable to XSS attacks.
CVE-2022-31025
LOWDiscourse is an open source platform for community discussion. Prior to version 2.8.4 on the `stable` branch and 2.9.0beta5 on the `beta` and `tests-passed` branches, inviting users on sites that use single sign-on could bypass the `must_approve_users` check and invites by staff are always approved automatically. The issue is patched in Discourse version 2.8.4 on the `stable` branch and version `2.9.0.beta5` on the `beta` and `tests-passed` branches. As a workaround, disable invites or increase `min_trust_level_to_allow_invite` to reduce the attack surface to more trusted users.
CVE-2022-24866
MEDIUMDiscourse Assign is a plugin for assigning users to a topic in Discourse, an open-source messaging platform. Prior to version 1.0.1, the UserBookmarkSerializer serialized the whole User / Group object, which leaked some private information. The data was only being serialized to people who could view assignment info, which is limited to staff by default. For the vast majority of sites, this data was only leaked to trusted staff member, but for sites with assign features enabled publicly, the data was accessible to more people than just staff. Version 1.0.1 contains a patch. There are currently no known workarounds.
CVE-2022-24850
MEDIUMDiscourse is an open source platform for community discussion. A category's group permissions settings can be viewed by anyone that has access to the category. As a result, a normal user is able to see whether a group has read/write permissions in the category even though the information should only be available to the users that can manage a category. This issue is patched in the latest stable, beta and tests-passed versions of Discourse. There are no workarounds for this problem.
CVE-2022-24824
MEDIUMDiscourse is an open source platform for community discussion. In affected versions an attacker can poison the cache for anonymous (i.e. not logged in) users, such that the users are shown the crawler view of the site instead of the HTML page. This can lead to a partial denial-of-service. This issue is patched in the latest stable, beta and tests-passed versions of Discourse. There are no known workarounds for this issue.
CVE-2022-24804
MEDIUMDiscourse is an open source platform for community discussion. In stable versions prior to 2.8.3 and beta versions prior 2.9.0.beta4 erroneously expose groups. When a group with restricted visibility has been used to set the permissions of a category, the name of the group is leaked to any user that is able to see the category. To workaround the problem, a site administrator can remove groups with restricted visibility from any category's permissions setting.
CVE-2022-24782
MEDIUMDiscourse is an open source discussion platform. Versions 2.8.2 and prior in the `stable` branch, 2.9.0.beta3 and prior in the `beta` branch, and 2.9.0.beta3 and prior in the `tests-passed` branch are vulnerable to a data leak. Users can request an export of their own activity. Sometimes, due to category settings, they may have category membership for a secure category. The name of this secure category is shown to the user in the export. The same thing occurs when the user's post has been moved to a secure category. A patch for this issue is available in the `main` branch of Discourse's GitHub repository and is anticipated to be part of future releases.
CVE-2022-23641
MEDIUMDiscourse is an open source discussion platform. In versions prior to 2.8.1 in the `stable` branch, 2.9.0.beta2 in the `beta` branch, and 2.9.0.beta2 in the `tests-passed` branch, users can trigger a Denial of Service attack by posting a streaming URL. Parsing Oneboxes in the background job trigger an infinite loop, which cause memory leaks. This issue is patched in version 2.8.1 of the `stable` branch, 2.9.0.beta2 of the `beta` branch, and 2.9.0.beta2 of the `tests-passed` branch. As a workaround, disable onebox in admin panel completely or specify allow list of domains that will be oneboxed.
CVE-2022-21677
MEDIUMDiscourse is an open source discussion platform. Discourse groups can be configured with varying visibility levels for the group as well as the group members. By default, a newly created group has its visibility set to public and the group's members visibility set to public as well. However, a group's visibility and the group's members visibility can be configured such that it is restricted to logged on users, members of the group or staff users. A vulnerability has been discovered in versions prior to 2.7.13 and 2.8.0.beta11 where the group advanced search option does not respect the group's visibility and members visibility level. As such, a group with restricted visibility or members visibility can be revealed through search with the right search option. This issue is patched in `stable` version 2.7.13, `beta` version 2.8.0.beta11, and `tests-passed` version 2.8.0.beta11 versions of Discourse. There are no workarounds aside from upgrading.
CVE-2022-21684
MEDIUMDiscourse is an open source discussion platform. Versions prior to 2.7.13 in `stable`, 2.8.0.beta11 in `beta`, and 2.8.0.beta11 in `tests-passed` allow some users to log in to a community before they should be able to do so. A user invited via email to a forum with `must_approve_users` enabled is going to be automatically logged in, bypassing the check that does not allow unapproved users to sign in. They will be able to do everything an approved user can do. If they logout, they cannot log back in. This issue is patched in the `stable` version 2.7.13, `beta` version 2.8.0.beta11, and `tests-passed` version 2.8.0.beta11. One may disable invites as a workaround. Administrators can increase `min_trust_level_to_allow_invite` to reduce the attack surface to more trusted users.
CVE-2022-21678
MEDIUMDiscourse is an open source discussion platform. Prior to version 2.8.0.beta11 in the `tests-passed` branch, version 2.8.0.beta11 in the `beta` branch, and version 2.7.13 in the `stable` branch, the bios of users who made their profiles private were still visible in the `<meta>` tags on their users' pages. The problem is patched in `tests-passed` version 2.8.0.beta11, `beta` version 2.8.0.beta11, and `stable` version 2.7.13 of Discourse.
CVE-2022-21642
MEDIUMDiscourse is an open source platform for community discussion. In affected versions when composing a message from topic the composer user suggestions reveals whisper participants. The issue has been patched in stable version 2.7.13 and beta version 2.8.0.beta11. There is no workaround for this issue and users are advised to upgrade.
CVE-2021-43850
MEDIUMDiscourse is an open source platform for community discussion. In affected versions admins users can trigger a Denial of Service attack via the `/message-bus/_diagnostics` path. The impact of this vulnerability is greater on multisite Discourse instances (where multiple forums are served from a single application server) where any admin user on any of the forums are able to visit the `/message-bus/_diagnostics` path. The problem has been patched. Please upgrade to 2.8.0.beta10 or 2.7.12. No workarounds for this issue exist.
CVE-2021-43840
MEDIUMmessage_bus is a messaging bus for Ruby processes and web clients. In versions prior to 3.3.7 users who deployed message bus with diagnostics features enabled (default off) are vulnerable to a path traversal bug, which could lead to disclosure of secret information on a machine if an unintended user were to gain access to the diagnostic route. The impact is also greater if there is no proxy for your web application as the number of steps up the directories is not bounded. For deployments which uses a proxy, the impact varies. For example, If a request goes through a proxy like Nginx with `merge_slashes` enabled, the number of steps up the directories that can be read is limited to 3 levels. This issue has been patched in version 3.3.7. Users unable to upgrade should ensure that MessageBus::Diagnostics is disabled.
CVE-2021-43827
MEDIUMdiscourse-footnote is a library providing footnotes for posts in Discourse. ### Impact When posting an inline footnote wrapped in `<a>` tags (e.g. `<a>^[footnote]</a>`, the resulting rendered HTML would include a nested `<a>`, which is stripped by Nokogiri because it is not valid. This then caused a javascript error on topic pages because we were looking for an `<a>` element inside the footnote reference span and getting its ID, and because it did not exist we got a null reference error in javascript. Users are advised to update to version 0.2. As a workaround editing offending posts from the rails console or the database console for self-hosters, or disabling the plugin in the admin panel can mitigate this issue.
CVE-2021-43793
MEDIUMDiscourse is an open source discussion platform. In affected versions a vulnerability in the Polls feature allowed users to vote multiple times in a single-option poll. The problem is patched in the latest tests-passed, beta and stable versions of Discourse
CVE-2021-43794
MEDIUMDiscourse is an open source discussion platform. In affected versions an attacker can poison the cache for anonymous (i.e. not logged in) users, such that the users are shown a JSON blob instead of the HTML page. This can lead to a partial denial-of-service. This issue is patched in the latest stable, beta and tests-passed versions of Discourse.
CVE-2021-43792
MEDIUMDiscourse is an open source discussion platform. In affected versions a vulnerability affects users of tag groups who use the "Tags are visible only to the following groups" feature. A tag group may only allow a certain group (e.g. staff) to view certain tags. Users who were tracking or watching the tags via /preferences/tags, then have their staff status revoked will still see notifications related to the tag, but will not see the tag on each topic. This issue has been patched in stable version 2.7.11. Users are advised to upgrade as soon as possible.
CVE-2021-41271
MEDIUMDiscourse is a platform for community discussion. In affected versions a maliciously crafted request could cause an error response to be cached by intermediate proxies. This could cause a loss of confidentiality for some content. This issue is patched in the latest stable, beta and tests-passed versions of Discourse.
CVE-2021-41263
HIGHrails_multisite provides multi-db support for Rails applications. In affected versions this vulnerability impacts any Rails applications using `rails_multisite` alongside Rails' signed/encrypted cookies. Depending on how the application makes use of these cookies, it may be possible for an attacker to re-use cookies on different 'sites' within a multi-site Rails application. The issue has been patched in v4 of the `rails_multisite` gem. Note that this upgrade will invalidate all previous signed/encrypted cookies. The impact of this invalidation will vary based on the application architecture.
CVE-2021-41163
CRITICALDiscourse is an open source platform for community discussion. In affected versions maliciously crafted requests could lead to remote code execution. This resulted from a lack of validation in subscribe_url values. This issue is patched in the latest stable, beta and tests-passed versions of Discourse. To workaround the issue without updating, requests with a path starting /webhooks/aws path could be blocked at an upstream proxy.
CVE-2021-41140
MEDIUMDiscourse-reactions is a plugin for the Discourse platform that allows user to add their reactions to the post. In affected versions reactions given by user to secure topics and private messages are visible. This issue is patched in version 0.2 of discourse-reaction. Users who are unable to update are advised to disable the Discourse-reactions plugin in admin panel.
CVE-2021-41095
MEDIUMDiscourse is an open source discussion platform. There is a cross-site scripting (XSS) vulnerability in versions 2.7.7 and earlier of the `stable` branch, versions 2.8.0.beta6 and earlier of the `beta` branch, and versions 2.8.0.beta6 and earlier of the `tests-passed` branch. Rendering of some error messages that contain user input can be susceptible to XSS attacks. This vulnerability only affects sites which have blocked watched words that contain HTML tags, modified or disabled Discourse's default Content Security Policy. This issue is patched in the latest `stable`, `beta` and `tests-passed` versions of Discourse. As a workaround, avoid modifying or disabling Discourse’s default Content Security Policy, and blocking watched words containing HTML tags.