VAITP Dataset

← Back to the dataset

CVE-2023-29374

LangChain 0.0.131's LLMMathChain is vulnerable to arbitrary code execution via Python exec method through prompt injection attacks

  • CVSS 9.8
  • CWE-74 Improper Neutralization of Special Elements in Output Used by a Downstream Component ('Injection')
  • Input Validation and Sanitization
  • Remote

In LangChain through 0.0.131, the LLMMathChain chain allows prompt injection attacks that can execute arbitrary code via the Python exec method.

CVSS base score
9.8
Published
2023-04-05
OWASP
A03 Injection
Orthogonal defect classification
Function
Code defect classification
Incorrect Functionality
Category
Input Validation and Sanitization
Subcategory
Command Injection
Accessibility scope
Remote
Impact
Arbitrary Code Execution
Affected component
LangChain
Fixed by upgrading
Yes

Solution

Update LangChain to version 0.0.132 or higher.

Vulnerable code sample

class LLMMathChain:
    def __init__(self, model):
        self.model = model

    def execute(self, prompt):
        # Directly execute the prompt using exec, which can be dangerous
        exec(prompt)

# Example usage
model = SomeLLMModel()  # Replace with actual model
chain = LLMMathChain(model)
chain.execute("print('This is an arbitrary code execution!')")  # Vulnerable input

Patched code sample

import re

class SafeLLMMathChain:
    def __init__(self, model):
        self.model = model

    def execute(self, prompt):
        # Sanitize the prompt to prevent code injection
        sanitized_prompt = self.sanitize_prompt(prompt)
        # Execute the model with the sanitized prompt
        return self.model.run(sanitized_prompt)

    def sanitize_prompt(self, prompt):
        # Remove any potentially dangerous code patterns
        safe_prompt = re.sub(r'[^a-zA-Z0-9\s\+\-\*/\(\)\.]+', '', prompt)
        return safe_prompt

# Example usage
model = SomeLLMModel()  # Replace with actual model
chain = SafeLLMMathChain(model)
result = chain.execute("2 + 2")  # Safe input

Published fix

Public sources link this CVE to the changes below. The code samples above were generated to illustrate the defect; these links lead to the code the project actually changed.

  • Patch LLMMathChain exec vulnerability (hwchase17/langchain)
    langchain/chains/llm_math/base.py, langchain/chains/llm_math/base.py · LLMMathChain._process_llm_result, langchain/chains/llm_math/prompt.py

Cite this entry

@misc{vaitp:cve202329374,
  title        = {{LangChain 0.0.131's LLMMathChain is vulnerable to arbitrary code execution via Python exec method through prompt injection attacks}},
  author       = {Bogaerts, Fr\'ed\'eric and Ivaki, Naghmeh and Fonseca, Jos\'e},
  year         = {2023},
  note         = {VAITP Python Vulnerability Dataset, entry CVE-2023-29374},
  howpublished = {\url{https://netpack.pt/vaitp/vulnerability/CVE-2023-29374/}}
}

Download this entry as JSON Download the whole dataset

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 ::

Before you use the VAITP AI models

Please read and accept the following disclaimer:

By choosing "I understand and accept", you acknowledge that you understand and agree to these terms.