Description

In JetBrains IntelliJ IDEA before 2026.3, 2026.2.1 command injection was possible in untrusted projects via Tomcat run configurations on WSL

Executive Summary

Technical Details

Mitigations

References

Additional Metadata

{
  "score": "7.8",
  "pubDate": "2026-11-06T17:17:22.097Z",
  "pubdate": "2026-11-06T17:17:22.097Z",
  "executiveSummary": "A command injection vulnerability has been identified in JetBrains IntelliJ IDEA, specifically affecting the handling of Tomcat run configurations when utilizing Windows Subsystem for Linux (WSL).\nThe vulnerability allows an unauthenticated attacker to execute arbitrary commands on the host system with the privileges of the IDE process.\nThis issue is triggered when a user opens a maliciously crafted, untrusted project within the IDE.\nThe impact includes full system compromise, unauthorized data access, and potential persistence mechanisms within the developer's environment.\nThe vulnerability exists in versions of IntelliJ IDEA prior to 2026.3 and 2026.2.1.\nExploitation requires the victim to trigger a specific run configuration, making it a social engineering-dependent vector where the primary risk involves executing code from unknown or untrusted sources.",
  "technicalDetails": "The vulnerability stems from improper sanitization of configuration parameters within the Tomcat run configuration component when interacting with the WSL environment.\nIn IntelliJ IDEA, Tomcat run configurations allow developers to define environment variables, VM options, and server parameters. When the IDE is configured to use a remote interpreter or build environment via WSL, these parameters are passed from the Windows host to the WSL distribution.\nThe root cause involves a failure to adequately validate or escape user-supplied inputs when constructing the command string that is executed inside the WSL shell context. Specifically, inputs provided within the 'VM options' or 'Environment variables' fields of the Tomcat configuration are incorrectly concatenated into the shell execution string without sufficient shell-metacharacter filtering.\nAn attacker can exploit this by crafting a project file (e.g., in .idea/runConfigurations/) that contains a malicious Tomcat configuration. When an unsuspecting user imports or opens the project and attempts to execute the Tomcat run configuration, the IDE constructs an execution command that includes the injected malicious payload.\nBecause the execution occurs within the WSL environment, the injected payload leverages the shell's ability to interpret command separators (such as ';', '&&', or '|'). This results in the execution of arbitrary commands under the shell session initiated by IntelliJ IDEA.\nThe attack flow is as follows: 1) Attacker distributes a project containing a backdoored .xml run configuration. 2) The victim opens the project in a vulnerable version of IntelliJ IDEA. 3) The IDE parses the configuration and prepares the execution command. 4) The malicious payload is concatenated into the command string. 5) The IDE executes the string via the WSL interop layer. 6) The operating system executes the attacker-defined commands within the WSL context.\nThis vulnerability is particularly dangerous because it bypasses standard UI-level protections, as the IDE treats these configurations as legitimate project settings. Successful exploitation grants the attacker the ability to execute code in the context of the WSL user, which often leads to lateral movement or further exploitation of the host system through WSL-Windows interoperability features.\nThe vulnerability affects all versions of IntelliJ IDEA prior to 2026.3 and 2026.2.1."
}