VAITP Dataset

← Back to the dataset

CVE-2026-19964

Jij-MCP-Server is vulnerable to remote code injection in PythonREPL.run.

  • CVSS 2.0
  • 74
  • Input Validation and Sanitization
  • Remote

A vulnerability was found in Jij-Inc Jij-MCP-Server 0.1.0. This affects the function PythonREPL.run of the file jij_mcp/python_repr.py of the component jm_check. The manipulation of the argument code results in code injection. It is possible to launch the attack remotely. The exploit has been made public and could be used. The project was informed of the problem early through an issue report but has not responded yet.

CWE
74
CVSS base score
2.0
Published
2026-08-17
OWASP
A03 Injection
Orthogonal defect classification
Checking
Code defect classification
Missing Check
Category
Input Validation and Sanitization
Subcategory
Command Injection
Accessibility scope
Remote
Impact
Arbitrary Code Execution
Fixed by upgrading
Yes

Solution

No patch is available. It is recommended to disable the `jm_check` component or discontinue use until a fix is released.

Vulnerable code sample

import sys
from io import StringIO
from typing import Any

# Placeholder for the external 'pb' module and its 'CheckResult' class
pb = type("pb", (), {"CheckResult": lambda **kwargs: None})


class PythonREPL:
    def __init__(self, globals: dict | None = None, locals: dict | None = None):
        self._globals = globals if globals is not None else {}
        self._locals = locals if locals is not None else {}

    def run(self, code: str) -> pb.CheckResult:
        old_stdout = sys.stdout
        sys.stdout = mystdout = StringIO()
        old_stderr = sys.stderr
        sys.stderr = mystderr = StringIO()
        try:
            # VULNERABLE: User-provided 'code' is executed with exec(), leading to code injection.
            exec(code, self._globals, self._locals)
        except Exception:
            pass
        finally:
            sys.stdout = old_stdout
            sys.stderr = old_stderr
        return pb.CheckResult(
            stdout=mystdout.getvalue(),
            stderr=mystderr.getvalue(),
        )

Patched code sample

import sys
from io import StringIO
from typing import Any
from RestrictedPython import compile_restricted, safe_builtins

# Placeholder for the external 'pb' module and its 'CheckResult' class
pb = type("pb", (), {"CheckResult": lambda **kwargs: None})


class PythonREPL:
    def __init__(self, globals: dict | None = None, locals: dict | None = None):
        self._globals = globals if globals is not None else {}
        self._locals = locals if locals is not None else {}

    def run(self, code: str) -> pb.CheckResult:
        old_stdout = sys.stdout
        sys.stdout = mystdout = StringIO()
        old_stderr = sys.stderr
        sys.stderr = mystderr = StringIO()
        try:
            # FIX: Compile the code in a restricted environment to prevent execution of malicious code.
            byte_code = compile_restricted(code, "<string>", "exec")
            safe_globals = {"__builtins__": safe_builtins}
            exec(byte_code, safe_globals, self._locals)
        except Exception:
            pass
        finally:
            sys.stdout = old_stdout
            sys.stderr = old_stderr
        return pb.CheckResult(
            stdout=mystdout.getvalue(),
            stderr=mystderr.getvalue(),
        )

Payload

import os; os.system('id')

Cite this entry

@misc{vaitp:cve202619964,
  title        = {{Jij-MCP-Server is vulnerable to remote code injection in PythonREPL.run.}},
  author       = {Bogaerts, Fr\'ed\'eric and Ivaki, Naghmeh and Fonseca, Jos\'e},
  year         = {2026},
  note         = {VAITP Python Vulnerability Dataset, entry CVE-2026-19964},
  howpublished = {\url{https://netpack.pt/vaitp/vulnerability/CVE-2026-19964/}}
}
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 ::