CyberRota Analysis
AI-GeneratedVersions of Langroid prior to 0.65.2 are critically vulnerable to a Sandbox Escape that allows for Remote Code Execution (RCE) through the `TableChatAgent` and `VectorStore` capabilities. The flaw arises from an incomplete sandboxing mechanism in Python's `eval()` function, which fails to restrict access to built-in functions, enabling attackers to execute arbitrary commands on the host system. Organizations utilizing Langroid for LLM-powered applications should prioritize upgrading to version 0.65.2 to mitigate this severe risk.
Public Exploit Signal
A public exploit, PoC, GitHub repository or Metasploit reference was detected for this CVE.
Note: these links are listed for security research and verification purposes only.
Original NVD Description
Langroid is a framework for building large-language-model-powered applications. Versions prior to 0.65.2 are vulnerable to a critical Sandbox Escape leading to Remote Code Execution (RCE) in its `TableChatAgent` and `VectorStore` capabilities. When these agents evaluate LLM-generated tool messages with `full_eval=True`, they attempt to sandbox the execution by explicitly setting `locals` to an empty dictionary `{}` inside Python's `eval()` function. However, this relies on an incomplete understanding of Python's execution model. Because `__builtins__` is not explicitly scrubbed from the `globals` dictionary mapping, Python implicitly injects all built-ins during execution, granting full access to functions like `__import__('os').system()`. Since `TableChatAgent.pandas_eval()` executes external LLM outputs natively, this bypass permits any attacker providing prompt payload to achieve unauthenticated RCE on the host system. Version 0.65.2 patches the issue.