CVE-2026-75062
CriticalCVSS 9.2Summary
Improper Neutralization of Directives in Dynamically Evaluated Code ('Eval Injection') in the default lf.query Python protocol in Google langfun versions prior to 0.1.2 allows remote unauthenticated attackers to execute arbitrary Python code in the context of the host application via crafted prompt inputs that cause the model to generate executable Python expressions evaluated without a sandbox.
Risk Assessment
An attacker can remotely execute arbitrary code in the application context, potentially leading to data compromise, system integrity violation, or privilege escalation.
Recommendation
Upgrade Google langfun to version 0.1.2 or later, which includes the fix for this vulnerability.
Original NVD description (English source)
Improper Neutralization of Directives in Dynamically Evaluated Code ('Eval Injection') in the default lf.query Python protocol in Google langfun versions prior to 0.1.2 allows remote unauthenticated attackers to execute arbitrary Python code in the context of the host application via crafted prompt inputs that cause the model to generate executable Python expressions evaluated without a sandbox.