Vm2 (<= 3.11.0) — CraftedSignal Threat Feedhttps://feed.craftedsignal.io/products/vm2--3.11.0/Trending threats, MITRE ATT&CK coverage, and detection metadata — refreshed continuously.Hugoenhello@craftedsignal.iohello@craftedsignal.ioSun, 28 Jan 2024 12:00:00 +0000vm2 NodeVM Nesting Bypass Allows Arbitrary Command Executionhttps://feed.craftedsignal.io/briefs/2024-01-28-vm2-sandbox-escape/Sun, 28 Jan 2024 12:00:00 +0000hello@craftedsignal.iohttps://feed.craftedsignal.io/briefs/2024-01-28-vm2-sandbox-escape/A vulnerability in vm2's NodeVM, when nesting is enabled, allows sandbox code to bypass require restrictions, enabling arbitrary OS command execution on the host.A critical vulnerability exists in vm2 versions 3.11.0 and below, specifically impacting the NodeVM when the nesting: true option is enabled. This flaw allows untrusted code running within the sandbox to bypass the intended require restrictions, even when require: false is explicitly set. By exploiting this bypass, malicious code can gain access to the vm2 module itself, create a new inner NodeVM with unrestricted permissions, and ultimately execute arbitrary OS commands on the host system. This can lead to complete compromise of applications relying on vm2 for secure code execution, affecting multi-tenant platforms, REPL services, and CI sandboxing environments. The vulnerability stems from how nesting: true overrides the require settings during module resolution, silently allowing access to vm2 even when it should be blocked.

Attack Chain

  1. The host application creates a NodeVM instance with nesting: true and require: false (or a restrictive require list) to sandbox untrusted code.
  2. The untrusted code within the sandbox calls require('vm2'). Due to the vulnerability, this succeeds despite the outer VM’s require restrictions.
  3. The sandbox code obtains the NodeVM constructor from the required vm2 module.
  4. The sandbox creates a new, inner NodeVM instance, specifying its own require configuration to include child_process.
  5. The inner NodeVM uses child_process.execSync() to execute an arbitrary OS command (e.g., id, whoami).
  6. The output of the executed command is converted to a string.
  7. The inner VM returns the command output to the outer VM.
  8. The outer VM returns the command output to the host application, effectively escaping the sandbox.

Impact

Successful exploitation allows attackers to execute arbitrary OS commands as the user running the host Node.js process. This gives the attacker the ability to read and write files, potentially exfiltrate sensitive information (secrets, API keys, etc.), move laterally within the network the host resides on, and establish persistent access to the compromised system. Any application employing vm2 with nesting: true to isolate untrusted code is vulnerable. This includes multi-tenant systems and CI/CD environments, posing a severe risk to infrastructure security. The vulnerability exists because developers expect require: false to provide a solid sandbox restriction, but enabling nesting: true silently overrides this expectation.

Recommendation

  • Immediately upgrade to a patched version of vm2 that addresses this vulnerability.
  • If upgrading is not immediately feasible, avoid using nesting: true in NodeVM configurations where untrusted code execution is involved.
  • Deploy the Sigma rule Detect vm2 Nesting Sandbox Escape via Child Process to identify potential exploitation attempts.
  • Enable process creation logging to support the detection rules.
  • Audit existing NodeVM configurations within your applications to identify instances where nesting: true is used in conjunction with restricted require settings.
  • Consider alternative sandboxing solutions that offer more robust module isolation if vm2 cannot be adequately secured.
]]>criticaladvisorysandbox-escapevm2code-execution