Latest Vulnerabilities with Public Exploits

NVIDIA Triton Inference Server for Windows and Linux contains a vulnerability in the Python backend, where an attacker could cause the shared memory limit to be exceeded by sending a very large request. A successful exploit of this vulnerability might lead to information disclosure.

Starlette is a lightweight ASGI framework/toolkit. Starting in version 0.39.0 and prior to version 0.49.1 , an unauthenticated attacker can send a crafted HTTP Range header that triggers quadratic-time processing in Starlette's FileResponse Range parsing/merging logic. This enables CPU exhaustion per request, causing denial‑of‑service for endpoints serving files (e.g., StaticFiles or any use of FileResponse). This vulnerability is fixed in 0.49.1.

Redis is an open source, in-memory database that persists on disk. Versions 8.2.1 and below allow an authenticated user to use a specially crafted Lua script to cause an integer overflow and potentially lead to remote code execution The problem exists in all versions of Redis with Lua scripting. This issue is fixed in version 8.2.2.

The Easy Timer plugin for WordPress is vulnerable to Remote Code Execution in all versions up to, and including, 4.2.1 via the plugin's shortcodes. This is due to insufficient restriction of shortcode attributes. This makes it possible for authenticated attackers, with Editor-level access and above, to execute code on the server.

Unrestricted Upload of File with Dangerous Type vulnerability in WebFactory AiBud WP aibuddy-openai-chatgpt allows Upload a Web Shell to a Web Server.This issue affects AiBud WP: from n/a through <= 1.9.

Pi-hole Admin Interface is a web interface for managing Pi-hole, a network-level advertisement and internet tracker blocking application. Pi-hole Admin Interface versions 6.2.1 and earlier are vulnerable to reflected cross-site scripting (XSS) via a malformed URL path. The 404 error page includes the requested path in the class attribute of the body tag without proper sanitization or escaping. An attacker can craft a URL containing an onload attribute that will execute arbitrary JavaScript code in the browser when a victim visits the malicious link. If an attacker sends a crafted pi-hole link to a victim and the victim visits it, attacker-controlled JavaScript code is executed in the browser of the victim. This has been patched in version 6.3.

An authentication bypass in the Palo Alto Networks PAN-OS software enables an unauthenticated attacker with network access to the management web interface to bypass the authentication otherwise required by the PAN-OS management web interface and invoke certain PHP scripts. While invoking these PHP scripts does not enable remote code execution, it can negatively impact integrity and confidentiality of PAN-OS. You can greatly reduce the risk of this issue by restricting access to the management web interface to only trusted internal IP addresses according to our recommended best practices deployment guidelines https://live.paloaltonetworks.com/t5/community-blogs/tips-amp-tricks-how-to-secure-the-management-access-of-your-palo/ba-p/464431 . This issue does not affect Cloud NGFW or Prisma Access software.

GALAYOU G2 cameras stream video output via RTSP streams. By default these streams are protected by randomly generated credentials. However these credentials are not required to access the stream. Changing these values does not change camera's behavior. The vendor did not respond in any way. Only version 11.100001.01.28 was tested, other versions might also be vulnerable.

A SQL injection vulnerability in the JS Jobs plugin versions 1.0.0-1.4.1 for Joomla allows low-privilege users to execute arbitrary SQL commands via the 'cvid' parameter in the employee application feature.

A SQL injection vulnerability in the JS Jobs plugin versions 1.1.5-1.4.2 for Joomla allows authenticated attackers (administrator) to execute arbitrary SQL commands via the 'fieldfor' parameter in the GDPR Field feature.

Formbricks is an open source qualtrics alternative. Prior to version 4.0.1, Formbricks is missing JWT signature verification. This vulnerability stems from a token validation routine that only decodes JWTs (jwt.decode) without verifying their signatures. Both the email verification token login path and the password reset server action use the same validator, which does not check the token’s signature, expiration, issuer, or audience. If an attacker learns the victim’s actual user.id, they can craft an arbitrary JWT with an alg: "none" header and use it to authenticate and reset the victim’s password. This issue has been patched in version 4.0.1.

The Keras Model.load_model function permits arbitrary code execution, even with safe_mode=True, through a manually constructed, malicious .keras archive. By altering the config.json file within the archive, an attacker can specify arbitrary Python modules and functions, along with their arguments, to be loaded and executed during model loading.

ssh in OpenSSH before 10.1 allows control characters in usernames that originate from certain possibly untrusted sources, potentially leading to code execution when a ProxyCommand is used. The untrusted sources are the command line and %-sequence expansion of a configuration file. (A configuration file that provides a complete literal username is not categorized as an untrusted source.)

Unity Runtime before 2025-10-02 on Android, Windows, macOS, and Linux allows argument injection that can result in loading of library code from an unintended location. If an application was built with a version of Unity Editor that had the vulnerable Unity Runtime code, then an adversary may be able to execute code on, and exfiltrate confidential information from, the machine on which that application is running. NOTE: product status is provided for Unity Editor because that is the information available from the Supplier. However, updating Unity Editor typically does not address the effects of the vulnerability; instead, it is necessary to rebuild and redeploy all affected applications.

A vulnerability has been found in Engeman Web up to 12.0.0.2. The affected element is an unknown function of the file /Login/RecoveryPass of the component Password Recovery Page. The manipulation of the argument LanguageCombobox as part of Cookie leads to sql injection. The attack is possible to be carried out remotely. The exploit has been disclosed to the public and may be used. Upgrading to version 12.0.0.3 is sufficient to fix this issue. Upgrading the affected component is advised. The vendor was contacted early about this disclosure but did not respond in any way.

RIOT-OS, an operating system that supports Internet of Things devices, has an ineffective size check implemented with `assert()` can lead to buffer overflow in versions up to and including 2025.04. Assertions are usually compiled out in production builds. If assertions are the only defense against untrusted inputs, the software may be exposed to attacks that utilize the lack of proper input checks. In the `l2filter_add()` function shown below, `addr_len` is checked using an assertion and is subsequently used as an argument in a `memcpy()` call. When assertions are disabled, there would be no size check for `addr_len`. As a consequence, if an attacker were to provide an `addr_len` value larger than `CONFIG_L2FILTER_ADDR_MAXLEN`, they can trigger a buffer overflow and write past the `list[i].addr` buffer. If the unchecked input is attacker-controlled, the impact of the buffer overflow can range from a denial of service to arbitrary code execution. Commit f6f7de4ccc107c018630e4c15500825caf02e1c2 contains a patch for the vulnerability.

OpenJPEG is an open-source JPEG 2000 codec. In OpenJPEG from 2.5.1 through 2.5.3, a call to opj_jp2_read_header may lead to OOB heap memory write when the data stream p_stream is too short and p_image is not initialized.

github.com/nwaples/rardecode versions <=2.1.1 fail to restrict the dictionary size when reading large RAR dictionary sizes, which allows an attacker to provide a specially crafted RAR file and cause Denial of Service via an Out Of Memory Crash.

In the Linux kernel, the following vulnerability has been resolved: xfrm: xfrm_alloc_spi shouldn't use 0 as SPI x->id.spi == 0 means "no SPI assigned", but since commit 94f39804d891 ("xfrm: Duplicate SPI Handling"), we now create states and add them to the byspi list with this value. __xfrm_state_delete doesn't remove those states from the byspi list, since they shouldn't be there, and this shows up as a UAF the next time we go through the byspi list.

A flaw was found in Undertow where malformed client requests can trigger server-side stream resets without triggering abuse counters. This issue, referred to as the "MadeYouReset" attack, allows malicious clients to induce excessive server workload by repeatedly causing server-side stream aborts. While not a protocol bug, this highlights a common implementation weakness that can be exploited to cause a denial of service (DoS).