An OS command injection vulnerability exists in the PowerShell and Cmd shell wrappers provided by the CliInvoke.Specializations package (the PowershellProcessInvoker/CmdProcessInvoker invokers, and the UsePowerShell/UseCmd middleware in v3 pre-release versions).
The wrappers re-run a caller-supplied target and arguments inside a shell command (pwsh -Command ... / cmd /c ...). In affected versions the wrapped command was delivered to the operating system as a single ProcessStartInfo.Arguments string. The OS command-line parser re-tokenizes that string before the shell parses it, so a double quote (") in the target or arguments breaks OS-level quoting and lets the wrapped shell reassemble a second, unintended command.
An attacker could exploit this to execute arbitrary commands with the privileges of the host process.
// patchesThe Specializations Packages now deliver the command via ProcessStartInfo.ArgumentList (natively where supported, and polyfilled in older TFMs), so that the operating system passes argv verbatim and only the shell parses the command once.
Upgrade to:
- 2.8.5 (2.8.x line)
- 2.9.4 (2.9.x line)
- 2.10.5 (2.10.x line)
- 3.0.0-beta.1 (3.x pre-release line)
No complete workaround is available. Until upgraded:
- Reject or strip " from any target path or argument passed to the PowerShell/Cmd wrappers. On 2.2.0 – 2.9.2 and 3.0.0-alpha.1 – alpha.4, also reject shell metacharacters (;, |, &, $, backtick, parentheses).
- Alternatively, bypass the wrappers for untrusted input and invoke the target process directly so that no second shell parse of the data occurs.