VAITP

vLLM, an inference and serving engine for large language models (LLMs), has a Regular Expression Denial of Service (ReDoS) vulnerability in the file `vllm/entrypoints/openai/tool_parsers/pythonic_tool_parser.py` of versions 0.6.4 up to but excluding 0.9.0. The root cause is the use of a highly complex and nested regular expression for tool call detection, which can be exploited by an attacker to cause severe performance degradation or make the service unavailable. The pattern contains multiple nested quantifiers, optional groups, and inner repetitions which make it vulnerable to catastrophic backtracking. Version 0.9.0 contains a patch for the issue.

Gradio is an open-source Python package that allows quick building of demos and web application for machine learning models, API, or any arbitrary Python function. Prior to version 5.31.0, an arbitrary file copy vulnerability in Gradio's flagging feature allows unauthenticated attackers to copy any readable file from the server's filesystem. While attackers can't read these copied files, they can cause DoS by copying large files (like /dev/urandom) to fill disk space. This issue has been patched in version 5.31.0.

A vulnerability classified as critical has been found in PyTorch 2.6.0. This affects the function torch.jit.script. The manipulation leads to memory corruption. It is possible to launch the attack on the local host. The exploit has been disclosed to the public and may be used.

A vulnerability was found in PyTorch 2.6.0. It has been rated as critical. Affected by this issue is the function torch.nn.utils.rnn.unpack_sequence. The manipulation leads to memory corruption. Attacking locally is a requirement. The exploit has been disclosed to the public and may be used.

When the Amazon Redshift Python Connector is configured with the BrowserAzureOAuth2CredentialsProvider plugin, the driver skips the SSL certificate validation step for the Identity Provider. An insecure connection could allow an actor to intercept the token exchange process and retrieve an access token. This issue has been addressed in driver version 2.1.7. Users should upgrade to address this issue and ensure any forked or derivative code is patched to incorporate the new fixes.

The Open edX Platform is a learning management platform. Prior to commit 6740e75c0fdc7ba095baf88e9f5e4f3e15cfd8ba, edxapp has no built-in protection against downloading the python_lib.zip asset from courses, which is a concern since it often contains custom grading code or answers to course problems. This potentially affects any course using custom Python-graded problem blocks. The openedx/configuration repo has had a patch since 2016 in the form of an nginx rule, but this was only intended as a temporary mitigation. As the configuration repo has been deprecated and we have not been able to locate any similar protection in Tutor, it is likely that most deployments have no protection against python_lib.zip being downloaded. The recommended mitigation, implemented in commit 6740e75c0fdc7ba095baf88e9f5e4f3e15cfd8ba, restricts python_lib.zip downloads to just the course team and site staff/superusers.

Langroid is a Python framework to build large language model (LLM)-powered applications. Prior to version 0.53.15, `LanceDocChatAgent` uses pandas eval() through `compute_from_docs()`. As a result, an attacker may be able to make the agent run malicious commands through `QueryPlan.dataframe_calc]`) compromising the host system. Langroid 0.53.15 sanitizes input to the affected function by default to tackle the most common attack vectors, and added several warnings about the risky behavior in the project documentation.

Langroid is a Python framework to build large language model (LLM)-powered applications. Prior to version 0.53.15, `TableChatAgent` uses `pandas eval()`. If fed by untrusted user input, like the case of a public-facing LLM application, it may be vulnerable to code injection. Langroid 0.53.15 sanitizes input to `TableChatAgent` by default to tackle the most common attack vectors, and added several warnings about the risky behavior in the project documentation.

setuptools is a package that allows users to download, build, install, upgrade, and uninstall Python packages. A path traversal vulnerability in `PackageIndex` is present in setuptools prior to version 78.1.1. An attacker would be allowed to write files to arbitrary locations on the filesystem with the permissions of the process running the Python code, which could escalate to remote code execution depending on the context. Version 78.1.1 fixes the issue.

Vyper is the Pythonic Programming Language for the Ethereum Virtual Machine. In versions up to and including 0.4.2rc1, `concat()` may skip evaluation of side effects when the length of an argument is zero. This is due to a fastpath in the implementation which skips evaluation of argument expressions when their length is zero. In practice, it would be very unusual in user code to construct zero-length bytestrings using an expression with side-effects, since zero-length bytestrings are typically constructed with the empty literal `b""`; the only way to construct an empty bytestring which has side effects would be with the ternary operator introduced in v0.3.8, e.g. `b"" if self.do_some_side_effect() else b""`. The fix is available in pull request 4644 and expected to be part of the 0.4.2 release. As a workaround, don't have side effects in expressions which construct zero-length bytestrings.