Описание
All versions of SuperAGI are vulnerable to Arbitrary Code Execution due to unsafe use of the ‘eval’ function. An attacker could induce the LLM output to exploit this vulnerability and gain arbitrary code execution on the SuperAGI application server.
All versions of SuperAGI are vulnerable to Arbitrary Code Execution due to unsafe use of the ‘eval’ function. An attacker could induce the LLM output to exploit this vulnerability and gain arbitrary code execution on the SuperAGI application server.
Ссылки
- https://nvd.nist.gov/vuln/detail/CVE-2024-21552
- https://github.com/TransformerOptimus/SuperAGI/blob/9361f0491716e56bd0c0ae2f3b49da201a18c58c/superagi/agent/output_handler.py#L149
- https://github.com/TransformerOptimus/SuperAGI/blob/9361f0491716e56bd0c0ae2f3b49da201a18c58c/superagi/agent/output_handler.py#L180
Связанные уязвимости
All versions of `SuperAGI` are vulnerable to Arbitrary Code Execution due to unsafe use of the ‘eval’ function. An attacker could induce the LLM output to exploit this vulnerability and gain arbitrary code execution on the SuperAGI application server.
Уязвимость функции eval() фреймворка для создания, управления и запуска автономных агентов на базе искуственного интеллекта SuperAGI, позволяющая нарушителю выполнить произвольный код и получить полный контроль над приложением