VAITP Dataset

← Back to the dataset

CVE-2024-6982

RCE via Python sandbox escape in lollms <= 9.8 using `eval()` and `_frozen_importlib`.

  • CVSS 8.4
  • CWE-94
  • Design Defects
  • Remote

A remote code execution vulnerability exists in the Calculate function of parisneo/lollms version 9.8. The vulnerability arises from the use of Python's `eval()` function to evaluate mathematical expressions within a Python sandbox that disables `__builtins__` and only allows functions from the `math` module. This sandbox can be bypassed by loading the `os` module using the `_frozen_importlib.BuiltinImporter` class, allowing an attacker to execute arbitrary commands on the server. The issue is fixed in version 9.10.

CVE
CVE-2024-6982
CWE
CWE-94
CVSS base score
8.4
Published
2025-03-20
OWASP
A03:2021-Injection
Orthogonal defect classification
Algorithm
Code defect classification
Incorrect Functionality
Category
Design Defects
Subcategory
Insecure Parsing or Deserialization
Accessibility scope
Remote
Impact
Arbitrary Code Execution
Affected component
parisneo/lol
Fixed by upgrading
Yes

Solution

Upgrade to version 9.10.

Vulnerable code sample

import math

def calculate(expression):
    """
    Calculates a mathematical expression using eval() with limited scope.
    """
    # Create a safe namespace for eval()
    safe_dict = {
        'math': math,
        '__builtins__': None  # Disable built-in functions
    }
    
    try:
        # Evaluate the expression within the safe namespace
        result = eval(expression, safe_dict)
        return result
    except Exception as e:
        return f"Error: {e}"

if __name__ == '__main__':
    while True:
        expression = input("Enter a mathematical expression (or 'exit'): ")
        if expression.lower() == 'exit':
            break
        result = calculate(expression)
        print("Result:", result)

Patched code sample

import os

import math
import ast
import operator as op

# Safe list of allowed operators
ALLOWED_OPERATORS = {
ast.Add: op.add,
ast.Sub: op.sub,
ast.Mult: op.mul,
ast.Div: op.truediv,
ast.Pow: op.pow,
ast.USub: op.neg
}


def safe_eval(expression):
    """
    Safely evaluates a mathematical expression using ast and a limited set of operators.
    """
    try:
        node = ast.parse(expression, mode='eval')

        def eval_(node):
            """Secure function that fixes the vulnerability."""
            if isinstance(node, ast.Num):
                return node.n
                elif isinstance(node, ast.BinOp):
                try:
                    return ALLOWED_OPERATORS[type(node.op)](eval_(node.left), eval_(node.right))
                    except KeyError:
                        raise ValueError("Operator not allowed")
                        elif isinstance(node, ast.UnaryOp):
                        try:
                            return ALLOWED_OPERATORS[type(node.op)](eval_(node.operand))
                            except KeyError:
                                raise ValueError("Operator not allowed")
                                elif isinstance(node, ast.Name):
                                if node.id in math.__dict__:
                                    return math.__dict__[node.id]
                                else:
                                    raise ValueError(f"Variable {node.id} not allowed")
                                    elif isinstance(node, ast.Call):
                #added support for basic functions to calculate things such as sin(x):
                                    if isinstance(node.func, ast.Name) and node.func.id in math.__dict__:
                                        func = math.__dict__[node.func.id]
                                        args = [eval_(arg) for arg in node.args]:
                                        return func(*args)
                                    else:
                                        raise ValueError(f"Function {node.func.id} not allowed")
                                    else:
                                        raise ValueError("Invalid expression")

                                        return eval_(node.body)
                                        except (SyntaxError, ValueError) as e:
                                            return f"Error: {e}"


# Example usage:
                                            if __name__ == '__main__':
                                                print(safe_eval("2 + 3 * 4"))  # Output: 14
                                                print(safe_eval("math.sqrt(25)")) # Output 5.0
                                                print(safe_eval("sin(0)"))  # Output: Error: Variable 'sin' not allowed
                                                print(safe_eval("math.sin(0)"))  # Output: 0.0
                                                print(safe_eval("2 ** 10")) # Output 1024
                                                print(safe_eval("().__class__.__bases__[0].__subclasses__()[59].__init__.__globals__['__builtins__']['eval']('__import__(\"os\").system(\"ls -l\")')")) # Output Error. Expression blocked from being processed.
                                                print(safe_eval("2 + os.system('ls -l')")) # Output: Error: Variable 'os' not allowed

Payload

_frozen_importlib.BuiltinImporter.load_module('os').system('id')

Cite this entry

@misc{vaitp:cve20246982,
  title        = {{RCE via Python sandbox escape in lollms <= 9.8 using `eval()` and `_frozen_importlib`.
}},
  author       = {Bogaerts, Fr\'ed\'eric and Ivaki, Naghmeh and Fonseca, Jos\'e},
  year         = {2025},
  note         = {VAITP Python Vulnerability Dataset, entry CVE-2024-6982},
  howpublished = {\url{https://netpack.pt/vaitp/vulnerability/CVE-2024-6982/}}
}
Introducing the "VAITP dataset": a specialized repository of Python vulnerabilities and patches, meticulously compiled for the use of the security research community. As Python's prominence grows, understanding and addressing potential security vulnerabilities become crucial. Crafted by and for the cybersecurity community, this dataset offers a valuable resource for researchers, analysts, and developers to analyze and mitigate the security risks associated with Python. Through the comprehensive exploration of vulnerabilities and corresponding patches, the VAITP dataset fosters a safer and more resilient Python ecosystem, encouraging collaborative advancements in programming security.

The supreme art of war is to subdue the enemy without fighting.

Sun Tzu – “The Art of War”

:: Shaping the future through research and ingenuity ::