VAITP Dataset

← Back to the dataset

CVE-2023-36095

Harrison Chase Langchain v0.0.194 has a critical arbitrary code execution via Python exec calls in PALChain functions from_math_prompt and from_colored_object_prompt

  • CVSS 9.8
  • CWE-94 Improper Control of Generation of Code ('Code Injection')
  • Input Validation and Sanitization
  • Remote

An issue in Harrison Chase langchain v.0.0.194 allows an attacker to execute arbitrary code via the python exec calls in the PALChain, affected functions include from_math_prompt and from_colored_object_prompt.

CVSS base score
9.8
Published
2023-08-05
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 Harrison Chase langchain to version 0.0.195 or higher.

Vulnerable code sample

class PALChain:
    def __init__(self):
        pass

    def from_math_prompt(self, prompt):
        # Vulnerable code using exec
        exec(f"result = {prompt}")
        return result

    def from_colored_object_prompt(self, prompt):
        # Vulnerable code using exec
        exec(f"result = {prompt}")
        return result

# Example usage
chain = PALChain()
# An attacker could input a malicious prompt
result = chain.from_math_prompt("os.system('echo Vulnerable')")
print(result)  # This could execute arbitrary code

Patched code sample

class SafePALChain:
    def __init__(self):
        pass

    def from_math_prompt(self, prompt):
        # Instead of using exec, use a safe evaluation method
        return self.safe_eval(prompt)

    def from_colored_object_prompt(self, prompt):
        # Instead of using exec, use a safe evaluation method
        return self.safe_eval(prompt)

    def safe_eval(self, expression):
        # Use a restricted environment for evaluation
        allowed_names = {"sqrt": __import__('math').sqrt}
        # Use eval with a limited scope to prevent arbitrary code execution
        return eval(expression, {"__builtins__": None}, allowed_names)

# Example usage
safe_chain = SafePALChain()
result = safe_chain.from_math_prompt("sqrt(16)")
print(result)  # Outputs: 4.0

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.

  • Mitigate issue #5872 (Prompt injection -> RCE in PAL chain) (#6003) (langchain-ai/langchain)
    langchain/chains/pal/base.py, langchain/chains/pal/base.py · PALChain.from_colored_object_prompt, langchain/chains/pal/base.py · PALChain.from_math_prompt, langchain/chains/pal/base.py · PALChain._call, langchain/utilities/python.py, langchain/utilities/python.py · PythonREPL.run
  • Some mitigations for RCE in PAL chain (#7870) (langchain-ai/langchain)
    langchain/chains/pal/base.py, langchain/chains/pal/base.py · PALChain.from_colored_object_prompt, langchain/chains/pal/base.py · PALChain.from_math_prompt, langchain/chains/pal/base.py · PALChain.raise_deprecation

Cite this entry

@misc{vaitp:cve202336095,
  title        = {{Harrison Chase Langchain v0.0.194 has a critical arbitrary code execution via Python exec calls in PALChain functions from_math_prompt and from_colored_object_prompt}},
  author       = {Bogaerts, Fr\'ed\'eric and Ivaki, Naghmeh and Fonseca, Jos\'e},
  year         = {2023},
  note         = {VAITP Python Vulnerability Dataset, entry CVE-2023-36095},
  howpublished = {\url{https://netpack.pt/vaitp/vulnerability/CVE-2023-36095/}}
}

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.