VAITP

There is a defect in the CPython standard library module “mimetypes” where on Windows the default list of known file locations are writable meaning other users can create invalid files to cause MemoryError to be raised on Python runtime startup or have file extensions be interpreted as the incorrect file type. This defect is caused by the default locations of Linux and macOS platforms (such as “/etc/mime.types”) also being used on Windows, where they are user-writable locations (“C:\etc\mime.types”). To work-around this issue a user can call mimetypes.init() with an empty list (“[]”) on Windows platforms to avoid using the default list of known file locations.

An authenticated attacker is able to use the Plugin Manager of the web interface of the Wattsense Bridge devices to upload malicious Python files to the device. This enables an attacker to gain remote root access to the device. An attacker needs a valid user account on the Wattsense web interface to be able to conduct this attack. This issue is fixed in recent firmware versions BSP >= 6.1.0.

PandasAI uses an interactive prompt function that is vulnerable to prompt injection and run arbitrary Python code that can lead to Remote Code Execution (RCE) instead of the intended explanation of the natural language processing by the LLM.

Wazuh is a free and open source platform used for threat prevention, detection, and response. Starting in version 4.4.0 and prior to version 4.9.1, an unsafe deserialization vulnerability allows for remote code execution on Wazuh servers. DistributedAPI parameters are a serialized as JSON and deserialized using `as_wazuh_object` (in `framework/wazuh/core/cluster/common.py`). If an attacker manages to inject an unsanitized dictionary in DAPI request/response, they can forge an unhandled exception (`__unhandled_exc__`) to evaluate arbitrary python code. The vulnerability can be triggered by anybody with API access (compromised dashboard or Wazuh servers in the cluster) or, in certain configurations, even by a compromised agent. Version 4.9.1 contains a fix.

vLLM is a high-throughput and memory-efficient inference and serving engine for LLMs. Maliciously constructed statements can lead to hash collisions, resulting in cache reuse, which can interfere with subsequent responses and cause unintended behavior. Prefix caching makes use of Python's built-in hash() function. As of Python 3.12, the behavior of hash(None) has changed to be a predictable constant value. This makes it more feasible that someone could try exploit hash collisions. The impact of a collision would be using cache that was generated using different content. Given knowledge of prompts in use and predictable hashing behavior, someone could intentionally populate the cache using a prompt known to collide with another prompt in use. This issue has been addressed in version 0.7.2 and all users are advised to upgrade. There are no known workarounds for this vulnerability.

A security vulnerability has been identified in the IBL Software Engineering Visual Weather and derived products (NAMIS, Aero Weather, Satellite Weather). The vulnerability is present in the Product Delivery Service (PDS) component in specific server configurations where the PDS pipeline utilizes the IPDS pipeline with Message Editor Output Filters enabled. A remote unauthenticated attacker can exploit this vulnerability to send unauthenticated requests to execute the IPDS pipeline with specially crafted Form Properties, enabling remote execution of arbitrary Python code. This vulnerability could lead to a full system compromise of the affected server, particularly if Visual Weather services are run under a privileged user account—contrary to the documented installation best practices. Upgrade to the patched versions 7.3.10 (or higher), 8.6.0 (or higher).

Django-Unicorn adds modern reactive component functionality to Django templates. Affected versions of Django-Unicorn are vulnerable to python class pollution vulnerability. The vulnerability arises from the core functionality `set_property_value`, which can be remotely triggered by users by crafting appropriate component requests and feeding in values of second and third parameter to the vulnerable function, leading to arbitrary changes to the python runtime status. With this finding at least five ways of vulnerability exploitation have been observed, stably resulting in Cross-Site Scripting (XSS), Denial of Service (DoS), and Authentication Bypass attacks in almost every Django-Unicorn-based application. This issue has been addressed in version 0.62.0 and all users are advised to upgrade. There are no known workarounds for this vulnerability.

The Python standard library functions `urllib.parse.urlsplit` and `urlparse` accepted domain names that included square brackets which isn't valid according to RFC 3986. Square brackets are only meant to be used as delimiters for specifying IPv6 and IPvFuture hosts in URLs. This could result in differential parsing across the Python URL parser and other specification-compliant URL parsers.

The Snowflake Connector for Python provides an interface for developing Python applications that can connect to Snowflake and perform all standard operations. Snowflake discovered and remediated a vulnerability in the Snowflake Connector for Python. The OCSP response cache uses pickle as the serialization format, potentially leading to local privilege escalation. This vulnerability affects versions 2.7.12 through 3.13.0. Snowflake fixed the issue in version 3.13.1.

The Snowflake Connector for Python provides an interface for developing Python applications that can connect to Snowflake and perform all standard operations. Snowflake discovered and remediated a vulnerability in the Snowflake Connector for Python. A function from the snowflake.connector.pandas_tools module is vulnerable to SQL injection. This vulnerability affects versions 2.2.5 through 3.13.0. Snowflake fixed the issue in version 3.13.1.