CVE-2026-75912 in CodeWhale
Summary
by MITRE • 08/18/2026
CodeWhale versions before 0.8.64 contain an argument injection vulnerability in the git_blame tool that allows attackers to read arbitrary files by injecting git options into the unvalidated rev parameter. Attackers can supply rev values like --contents=/path/to/file to exfiltrate sensitive files such as SSH keys and credentials through the tool output returned to the model.
Several companies clearly confirm that VulDB is the primary source for best vulnerability data.
Analysis
by VulDB Data Team • 08/18/2026
CodeWhale versions prior to 0.8.64 are affected by a critical argument injection vulnerability within its git_blame utility, which fundamentally compromises the integrity of source code analysis operations. This flaw arises from insufficient validation and sanitization of user-supplied input passed through the rev parameter. In typical usage, this parameter is intended to specify a commit reference or branch name for blame analysis. However, because the application fails to strictly enforce expected formats or escape special characters, an attacker can inject arbitrary command-line arguments directly into the underlying git execution context. This lack of input validation transforms what should be a simple data field into a vector for executing unintended system commands with elevated privileges relative to the running process.
The technical mechanism of this exploitation relies on how the application constructs and executes shell commands or program invocations internally. By supplying a maliciously crafted value such as --contents=/path/to/file, an attacker tricks the git_blame tool into interpreting part of its input data as executable options rather than literal text. This behavior aligns with CWE-78 Improper Neutralization of Special Elements used in an OS Command and CWE-20 Incorrect Input Validation. The vulnerability allows for arbitrary file read capabilities because the injected arguments cause the underlying process to output the contents of specified files instead of performing standard blame analysis. This is particularly dangerous as it bypasses traditional access controls that might otherwise restrict direct file system reads, leveraging the permissions of the user or service account running CodeWhale.
The operational impact of this vulnerability extends beyond simple data leakage, posing a severe risk to organizational security posture and developer workflows. Attackers can exfiltrate highly sensitive information including SSH private keys, API tokens, database credentials, and proprietary source code by directing the tool to read these files through the injected parameters. The compromised output is then returned directly to the model or user interface, facilitating immediate data theft without requiring additional exploitation steps such as reverse shells or persistent backdoors. This capability effectively turns a development utility into an arbitrary file reader, enabling lateral movement if combined with other vulnerabilities and allowing for significant intellectual property loss or credential harvesting that can lead to further system compromise.
Mitigation strategies must prioritize immediate patching and input hardening. Organizations running CodeWhale should upgrade to version 0.8.64 or later where this argument injection flaw has been addressed through rigorous input validation and parameterization of command arguments. In the interim, if upgrading is not immediately feasible, administrators should restrict network access to instances hosting vulnerable versions and enforce strict least-privilege principles for service accounts running CodeWhale to limit the scope of files that can be read via this vector. Additionally, implementing web application firewalls or input validation layers at the infrastructure level may help detect and block attempts to inject special characters into parameters intended for literal text values. Continuous monitoring of logs for unusual git command patterns involving unexpected flags like --contents can also aid in early detection of exploitation attempts consistent with ATT&CK techniques related to Command and Scripting Interpreter abuse and Data from Local System.