Xstream API versions up to 1.4.6 and version 1.4.10, if the security framework has not been initialized, may allow a remote attacker to run arbitrary shell commands by manipulating the processed input stream when unmarshaling XML or any supported format. e.g. JSON.

References

• https://nvd.nist.gov/vuln/detail/CVE-2013-7285
• http://blog.diniscruz.com/2013/12/xstream-remote-code-execution-exploit.html
• http://seclists.org/oss-sec/2014/q1/69
• https://www.mail-archive.com/user@xstream.codehaus.org/msg00604.html
• https://www.mail-archive.com/user@xstream.codehaus.org/msg00607.html
• https://x-stream.github.io/CVE-2013-7285.html
• https://lists.apache.org/thread.html/6d3d34adcf3dfc48e36342aa1f18ce3c20bb8e4c458a97508d5bfed1@%3Cissues.activemq.apache.org%3E
• https://lists.apache.org/thread.html/dcf8599b80e43a6b60482607adb76c64672772dc2d9209ae2170f369@%3Cissues.activemq.apache.org%3E
• https://www.oracle.com/security-alerts/cpuoct2020.html
• http://web.archive.org/web/20140204133306/http://blog.diniscruz.com/2013/12/xstream-remote-code-execution-exploit.html
• x-stream/xstream@6344867