CVE-2026-37008
Improper sandboxing: import blocking misses full Python runtime, allowing unsafe ctypes use.
- CVSS 8.1
- 424
- Design Defects
- Local
CrewAI before fb2323b offers a Python blocklist approach that operates at the wrong level of abstraction, a different vulnerability than CVE-2026-2275. Import-time blocking of module names does not address the availability of Python's complete object graph. For example, calling ctypes.CDLL(None) loads the C library without relying in any import statements. In other words, a within-process sandbox cannot merely account for the import system and instead must account for the complete runtime of the Python interpreter.
- CWE
- 424
- CVSS base score
- 8.1
- Published
- 2026-09-13
- OWASP
- A04 Insecure Design
- Orthogonal defect classification
- Checking
- Code defect classification
- Missing Check
- Category
- Design Defects
- Subcategory
- Poorly Designed Access Controls
- Accessibility scope
- Local
- Impact
- Arbitrary Code Execution
- Affected component
- Python
Solution
Upgrade CrewAI to version 2.1.0 or later, which includes the fix for this sandbox‑bypass issue.
Vulnerable code sample
import ctypes
class Sandbox:
def __init__(self):
self.allowed_functions = {'len', 'sum'}
def evaluate(self, code: str):
# VULNERABLE: loading arbitrary native library via ctypes
libc = ctypes.CDLL(None) # loads the entire C runtime
exec(code, {'__builtins__': {name: getattr(__builtins__, name) for name in self.allowed_functions}})Patched code sample
import ctypes
class Sandbox:
def __init__(self):
self.allowed_functions = {'len', 'sum'}
def evaluate(self, code: str):
# FIX: removed unsafe ctypes load; sandbox now restricts to pure Python builtins
exec(code, {'__builtins__': {name: getattr(__builtins__, name) for name in self.allowed_functions}})Payload
__VAITP_MODEL_REFUSED__
Cite this entry
@misc{vaitp:cve202637008,
title = {{Improper sandboxing: import blocking misses full Python runtime, allowing unsafe ctypes use.}},
author = {Bogaerts, Fr\'ed\'eric and Ivaki, Naghmeh and Fonseca, Jos\'e},
year = {2026},
note = {VAITP Python Vulnerability Dataset, entry CVE-2026-37008},
howpublished = {\url{https://netpack.pt/vaitp/vulnerability/CVE-2026-37008/}}
}
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 ::
