Vulnerabilities

Several rest inline action resources of Atlassian Activity Streams before version 6.3.0 allows remote authenticated attackers to watch any Confluence page & receive notifications when comments are added to the watched page, and vote & watch JIRA issues that they do not have access to, although they will not receive notifications for the issue, via missing permission checks.

Sourcetree for Windows had several argument and command injection bugs in Mercurial and Git repository handling. An attacker with permission to commit to a repository linked in Sourcetree for Windows is able to exploit this issue to gain code execution on the system. From version 0.8.4b of Sourcetree for Windows, this vulnerability can be triggered from a webpage through the use of the Sourcetree URI handler. Versions of Sourcetree for Windows starting with 0.5.1.0 before version 2.4.7.0 are affected by this vulnerability

Sourcetree for macOS had several argument and command injection bugs in Mercurial and Git repository handling. An attacker with permission to commit to a repository linked in Sourcetree for macOS is able to exploit this issue to gain code execution on the system. From version 1.4.0 of Sourcetree for macOS, this vulnerability can be triggered from a webpage through the use of the Sourcetree URI handler. Versions of Sourcetree for macOS starting with 1.0b2 before version 2.7.0 are affected by this vulnerability.

The PieChart gadget in Atlassian Jira before version 7.5.3 allows remote attackers to inject arbitrary HTML or JavaScript via a cross site scripting (XSS) vulnerability through the name of a project or filter.

The Jira-importers-plugin in Atlassian Jira before version 7.6.1 allows remote attackers to create new projects and abort an executing external system import via various Cross-site request forgery (CSRF) vulnerabilities.

The Trello importer in Atlassian Jira before version 7.6.1 allows remote attackers to access the content of internal network resources via a Server Side Request Forgery (SSRF). When running in an environment like Amazon EC2, this flaw maybe used to access to a metadata resource that provides access credentials and other potentially confidential information.

The printable searchrequest issue resource in Atlassian Jira before version 7.2.12 and from version 7.3.0 before 7.6.1 allows remote attackers to inject arbitrary HTML or JavaScript via a cross site scripting (XSS) vulnerability in the jqlQuery query parameter.

The issue search resource in Atlassian Jira before version 7.4.2 allows remote attackers to inject arbitrary HTML or JavaScript via a cross site scripting (XSS) vulnerability in the orderby parameter.

The IncomingMailServers resource in Atlassian Jira before version 7.6.2 allows remote attackers to modify the "incoming mail" whitelist setting via a Cross-site request forgery (CSRF) vulnerability.

Bamboo did not check that the name of a branch in a Mercurial repository contained argument parameters. An attacker who has permission to create a repository in Bamboo, edit an existing plan that has a non-linked Mercurialrepository, create or edit a plan when there is at least one linked Mercurial repository that the attacker has permission to use, or commit to a Mercurial repository used by a Bamboo plan which has branch detection enabled can execute code of their choice on systems that run a vulnerable version of Bamboo Server. Versions of Bamboo starting with 2.7.0 before 6.1.6 (the fixed version for 6.1.x) and from 6.2.0 before 6.2.5 (the fixed version for 6.2.x) are affected by this vulnerability.

It was possible for double OGNL evaluation in FreeMarker templates through Struts FreeMarker tags to occur. An attacker who has restricted administration rights to Bamboo or who hosts a website that a Bamboo administrator visits, is able to exploit this vulnerability to execute Java code of their choice on systems that run a vulnerable version of Bamboo. All versions of Bamboo before 6.1.6 (the fixed version for 6.1.x) and from 6.2.0 before 6.2.5 (the fixed version for 6.2.x) are affected by this vulnerability.

The RSS Feed macro in Atlassian Confluence before version 6.5.2 allows remote attackers to inject arbitrary HTML or JavaScript via cross site scripting (XSS) vulnerabilities in various rss properties which were used as links without restriction on their scheme.

It is possible to bypass the bitbucket auto-unapprove plugin via minimal brute-force because it is relying on asynchronous events on the back-end. This allows an attacker to merge any code into unsuspecting repositories. This affects all versions of the auto-unapprove plugin, however since the auto-unapprove plugin is not bundled with Bitbucket Server it does not affect any particular version of Bitbucket.

Atlassian Fisheye and Crucible versions less than 4.4.3 and version 4.5.0 are vulnerable to argument injection through filenames in Mercurial repositories, allowing attackers to execute arbitrary code on a system running the impacted software.

A Server Side Request Forgery (SSRF) vulnerability could lead to remote code execution for authenticated administrators. This issue was introduced in version 2.2.0 of Hipchat Server and version 3.0.0 of Hipchat Data Center. Versions of Hipchat Server starting with 2.2.0 and before 2.2.6 are affected by this vulnerability. Versions of Hipchat Data Center starting with 3.0.0 and before 3.1.0 are affected.

The Hipchat for Mac desktop client is vulnerable to client-side remote code execution via video call link parsing. Hipchat for Mac desktop clients at or above version 4.0 and before version 4.30 are affected by this vulnerability.

Bamboo before 6.0.5, 6.1.x before 6.1.4, and 6.2.x before 6.2.1 had a REST endpoint that parsed a YAML file and did not sufficiently restrict which classes could be loaded. An attacker who can log in to Bamboo as a user is able to exploit this vulnerability to execute Java code of their choice on systems that have vulnerable versions of Bamboo.

The administration user deletion resource in Atlassian Fisheye and Crucible before version 4.4.2 allows remote attackers to inject arbitrary HTML or JavaScript via a cross site scripting (XSS) vulnerability in the uname parameter.

Various resources in Atlassian Fisheye and Crucible before version 4.4.2 allow remote attackers to inject arbitrary HTML or JavaScript via a cross site scripting (XSS) vulnerability in the dialog parameter.

The MultiPathResource class in Atlassian Fisheye and Crucible, before version 4.4.1 allows anonymous remote attackers to read arbitrary files via a path traversal vulnerability when Fisheye or Crucible is running on the Microsoft Windows operating system.

Various resources in Atlassian Fisheye and Crucible before version 4.4.1 allow remote attackers to inject arbitrary HTML or JavaScript via a cross site scripting (XSS) vulnerability through the name of a repository or review file.

The review file upload resource in Atlassian Crucible before version 4.4.1 allows remote attackers to inject arbitrary HTML or JavaScript via a cross site scripting (XSS) vulnerability through the charset of a previously uploaded file.

The repository changelog resource in Atlassian Fisheye before version 4.4.1 allows remote attackers to inject arbitrary HTML or JavaScript via a cross site scripting (XSS) vulnerability through the start date and end date parameters.

The review dashboard resource in Atlassian Crucible from version 4.1.0 before version 4.4.1 allows remote attackers to inject arbitrary HTML or JavaScript via a cross site scripting (XSS) vulnerability in the review filter title parameter.

The mostActiveCommitters.do resource in Atlassian Fisheye and Crucible, before version 4.4.1 allows anonymous remote attackers to access sensitive information, for example email addresses of committers, as it lacked permission checks.

The IconUriServlet of the Atlassian OAuth Plugin from version 1.3.0 before version 1.9.12 and from version 2.0.0 before version 2.0.4 allows remote attackers to access the content of internal network resources and/or perform an XSS attack via Server Side Request Forgery (SSRF).

Atlassian Confluence starting with 4.3.0 before 6.2.1 did not check if a user had permission to view a page when creating a workbox notification about new comments. An attacker who can login to Confluence could receive workbox notifications, which contain the content of comments, for comments added to a page after they started watching it even if they do not have permission to view the page itself.

Atlassian Bamboo 5.x before 5.15.7 and 6.x before 6.0.1 did not correctly check if a user creating a deployment project had the edit permission and therefore the rights to do so. An attacker who can login to Bamboo as a user without the edit permission for deployment projects is able to use this vulnerability, provided there is an existing plan with a green build, to create a deployment project and execute arbitrary code on an available Bamboo Agent. By default a local agent is enabled; this means that code execution can occur on the system hosting Bamboo as the user running Bamboo.