VAITP Dataset

← Back to the dataset

CVE-2026-127208

Code injection via crafted x‑python‑type in JSON Schema leads to arbitrary exec in datamodel‑code‑generator.

  • CVSS 8.6
  • CWE-94
  • Input Validation and Sanitization
  • Remote

datamodel-code-generator before 0.84.0 contains a code injection vulnerability that allows attackers to achieve arbitrary code execution by supplying a malicious JSON Schema with a crafted x-python-type annotation containing a string ast.Constant node in a subscript slice position, which bypasses validation in parse_python_type_annotation and _python_type_expr_from_ast. Attackers can embed malicious Python expressions such as __import__ calls within quoted string literals in the generated source code, which Pydantic v2 evaluates as a ForwardRef via eval() during model_rebuild(), resulting in arbitrary code execution on developer workstations, CI/CD pipelines, or applications that generate and import the produced models at runtime.

CVSS base score
8.6
Published
2026-11-12
OWASP
A03 Injection
Orthogonal defect classification
Checking
Code defect classification
Missing Check
Category
Input Validation and Sanitization
Subcategory
Insecure Parsing or Deserialization
Accessibility scope
Remote
Impact
Arbitrary Code Execution
Affected component
datamodel-code-generator
Fixed by upgrading
Yes

Solution

Upgrade datamodel‑code‑generator to version 0.84.0 or later.

Vulnerable code sample

import ast
from typing import Any, Dict

def generate_model(schema: Dict[str, Any]) -> str:
    """
    Very simplified version of datamodel-code-generator's handling of
    the custom x-python-type annotation.
    """
    fields = []
    for prop, definition in schema.get("properties", {}).items():
        # VULNERABLE: directly embed unvalidated annotation string into AST
        ann = definition.get("x-python-type", "str")
        # Build a fake subscript expression like List["__import__('os').system('rm -rf /')"]
        expr = ast.parse(f"typing.{ann}", mode="eval").body
        # Convert AST back to source (simulating code generation)
        annotation = ast.unparse(expr)
        fields.append(f"    {prop}: {annotation}")
    return "\n".join(["class Model:", *fields])

Patched code sample

import ast
from typing import Any, Dict

def generate_model(schema: Dict[str, Any]) -> str:
    """
    Hardened version that validates the AST of x-python-type before use.
    """
    fields = []
    for prop, definition in schema.get("properties", {}).items():
        # FIX: validate AST to allow only Name or Subscript with Name bases
        ann = definition.get("x-python-type", "str")
        expr = ast.parse(f"typing.{ann}", mode="eval").body
        if not _is_safe_type_expr(expr):
            raise ValueError("Unsafe type annotation")
        annotation = ast.unparse(expr)
        fields.append(f"    {prop}: {annotation}")
    return "\n".join(["class Model:", *fields])

def _is_safe_type_expr(node: ast.AST) -> bool:
    """
    Allow only simple Name nodes and Subscript where the slice is a Name.
    """
    if isinstance(node, ast.Name):
        return True
    if isinstance(node, ast.Subscript):
        # base must be Name (e.g., List) and slice must be Name (e.g., int)
        return isinstance(node.value, ast.Name) and isinstance(node.slice, ast.Name)
    return False

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.

  • Merge commit from fork (datamodel-code-generator/datamodel-code-generator)
    src/datamodel_code_generator/_python_type_annotation_codec.py, src/datamodel_code_generator/_python_type_annotation_codec.py · _python_type_expr_from_ast

Cite this entry

@misc{vaitp:cve2026127208,
  title        = {{Code injection via crafted x‑python‑type in JSON Schema leads to arbitrary exec in datamodel‑code‑generator.}},
  author       = {Bogaerts, Fr\'ed\'eric and Ivaki, Naghmeh and Fonseca, Jos\'e},
  year         = {2026},
  note         = {VAITP Python Vulnerability Dataset, entry CVE-2026-127208},
  howpublished = {\url{https://netpack.pt/vaitp/vulnerability/CVE-2026-127208/}}
}

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.