| CVE |
Vendors |
Products |
Updated |
CVSS v3.1 |
| SillyTavern is a locally installed user interface that allows users to interact with text generation large language models, image generation engines, and text-to-speech voice models. In versions prior to 1.13.4, the web user interface for SillyTavern is susceptible to DNS rebinding, allowing attackers to perform actions like install malicious extensions, read chats, inject arbitrary HTML for phishing attacks, etc. The vulnerability has been patched in the version 1.13.4 by introducing a server configuration setting that enables a validation of host names in inbound HTTP requests according to the provided list of allowed hosts: `hostWhitelist.enabled` in config.yaml file or `SILLYTAVERN_HOSTWHITELIST_ENABLED` environment variable. While the setting is disabled by default to honor a wide variety of existing user configurations and maintain backwards compatibility, existing and new users are encouraged to review their server configurations and apply necessary changes to their setup, especially if hosting over the local network while not using SSL. |
| vLLM is an inference and serving engine for large language models (LLMs). Before version 0.11.0rc2, the API key support in vLLM performs validation using a method that was vulnerable to a timing attack. API key validation uses a string comparison that takes longer the more characters the provided API key gets correct. Data analysis across many attempts could allow an attacker to determine when it finds the next correct character in the key sequence. Deployments relying on vLLM's built-in API key validation are vulnerable to authentication bypass using this technique. Version 0.11.0rc2 fixes the issue. |
| Components of the YoSmart YoLink ecosystem through 2025-10-02 leverage unencrypted MQTT to communicate over the internet. An attacker with the ability to monitor network traffic could therefore obtain sensitive information or tamper with the traffic to control affected devices. This affects YoLink Hub 0382, YoLink Mobile Application 1.40.41, and YoLink MQTT Broker. |
| Cross Site Request Forgery (CSRF) vulnerability in EndRun Technologies Sonoma D12 Network Time Server (GPS) F/W 6010-0071-000 Ver 4.00 allows attackers to execute arbitrary code, cause a denial of service, gain escalated privileges, and gain sensitive information. |
| Cross Site Scripting (XSS) vulnerability in EndRun Technologies Sonoma D12 Network Time Server (GPS) F/W 6010-0071-000 Ver 4.00 allows attackers to gain sensitive information. |
| Cross Site Scripting (XSS) vulnerability in EndRun Technologies Sonoma D12 Network Time Server (GPS) F/W 6010-0071-000 Ver 4.00 allows attackers to gain sensitive information, and possibly other unspecified impacts. |
| OS Command Injection vulnerability in EndRun Technologies Sonoma D12 Network Time Server (GPS) F/W 6010-0071-000 Ver 4.00 allows attackers to gain sensitive information, and possibly other unspecified impacts. |
| OS Command Injection vulnerability in EndRun Technologies Sonoma D12 Network Time Server (GPS) F/W 6010-0071-000 Ver 4.00 allows attackers to execute arbitrary code, cause a denial of service, gain escalated privileges, gain sensitive information, and possibly other unspecified impacts. |
| Flowise is a drag & drop user interface to build a customized large language model flow. A file upload vulnerability in version 3.0.7 of FlowiseAI allows authenticated users to upload arbitrary files without proper validation. This enables attackers to persistently store malicious Node.js web shells on the server, potentially leading to Remote Code Execution (RCE). The system fails to validate file extensions, MIME types, or file content during uploads. As a result, malicious scripts such as Node.js-based web shells can be uploaded and stored persistently on the server. These shells expose HTTP endpoints capable of executing arbitrary commands if triggered. The uploaded shell does not automatically execute, but its presence allows future exploitation via administrator error or chained vulnerabilities. This presents a high-severity threat to system integrity and confidentiality. As of time of publication, no known patched versions are available. |
| Emlog is an open source website building system. A cross-site scripting (XSS) vulnerability in emlog up to and including version 2.5.22 allows authenticated remote attackers to inject arbitrary web script or HTML via the file upload functionality. As an authenticated user it is possible to upload .svg file that contains JavaScript code that is later being executed. Commit 052f9c4226b2c0014bcd857fec47677340b185b1 fixes the issue. |
| Rack is a modular Ruby web server interface. In versions prior to 2.2.19, 3.1.17, and 3.2.2, `Rack::Multipart::Parser` buffers the entire multipart preamble (bytes before the first boundary) in memory without any size limit. A client can send a large preamble followed by a valid boundary, causing significant memory use and potential process termination due to out-of-memory (OOM) conditions. Remote attackers can trigger large transient memory spikes by including a long preamble in multipart/form-data requests. The impact scales with allowed request sizes and concurrency, potentially causing worker crashes or severe slowdown due to garbage collection. Versions 2.2.19, 3.1.17, and 3.2.2 enforce a preamble size limit (e.g., 16 KiB) or discard preamble data entirely. Workarounds include limiting total request body size at the proxy or web server level and monitoring memory and set per-process limits to prevent OOM conditions. |
| Rack is a modular Ruby web server interface. In versions prior to 2.2.19, 3.1.17, and 3.2.2, ``Rack::Multipart::Parser` stores non-file form fields (parts without a `filename`) entirely in memory as Ruby `String` objects. A single large text field in a multipart/form-data request (hundreds of megabytes or more) can consume equivalent process memory, potentially leading to out-of-memory (OOM) conditions and denial of service (DoS). Attackers can send large non-file fields to trigger excessive memory usage. Impact scales with request size and concurrency, potentially leading to worker crashes or severe garbage-collection overhead. All Rack applications processing multipart form submissions are affected. Versions 2.2.19, 3.1.17, and 3.2.2 enforce a reasonable size cap for non-file fields (e.g., 2 MiB). Workarounds include restricting maximum request body size at the web-server or proxy layer (e.g., Nginx `client_max_body_size`) and validating and rejecting unusually large form fields at the application level. |
| Rack is a modular Ruby web server interface. In versions prior to 2.2.19, 3.1.17, and 3.2.2, `Rack::Multipart::Parser` can accumulate unbounded data when a multipart part’s header block never terminates with the required blank line (`CRLFCRLF`). The parser keeps appending incoming bytes to memory without a size cap, allowing a remote attacker to exhaust memory and cause a denial of service (DoS). Attackers can send incomplete multipart headers to trigger high memory use, leading to process termination (OOM) or severe slowdown. The effect scales with request size limits and concurrency. All applications handling multipart uploads may be affected. Versions 2.2.19, 3.1.17, and 3.2.2 cap per-part header size (e.g., 64 KiB). As a workaround, restrict maximum request sizes at the proxy or web server layer (e.g., Nginx `client_max_body_size`). |
| Flag Forge is a Capture The Flag (CTF) platform. Starting in version 2.0.0 and prior to version 2.3.2, the `/api/admin/badge-templates` (GET) and `/api/admin/badge-templates/create` (POST) endpoints previously allowed access without authentication or authorization. This could have enabled unauthorized users to retrieve all badge templates and sensitive metadata (createdBy, createdAt, updatedAt) and/or create arbitrary badge templates in the database. This could lead to data exposure, database pollution, or abuse of the badge system. The issue has been fixed in FlagForge v2.3.2. GET, POST, UPDATE, and DELETE endpoints now require authentication. Authorization checks ensure only admins can access and modify badge templates. No reliable workarounds are available. |
| Akka.NET is a .NET port of the Akka project from the Scala / Java community. In all versions of Akka.Remote from v1.2.0 to v1.5.51, TLS could be enabled via our `akka.remote.dot-netty.tcp` transport and this would correctly enforce private key validation on the server-side of inbound connections. Akka.Remote, however, never asked the outbound-connecting client to present ITS certificate - therefore it's possible for untrusted parties to connect to a private key'd Akka.NET cluster and begin communicating with it without any certificate. The issue here is that for certificate-based authentication to work properly, ensuring that all members of the Akka.Remote network are secured with the same private key, Akka.Remote needed to implement mutual TLS. This was not the case before Akka.NET v1.5.52. Those who run Akka.NET inside a private network that they fully control or who were never using TLS in the first place are now affected by the bug. However, those who use TLS to secure their networks must upgrade to Akka.NET V1.5.52 or later. One patch forces "fail fast" semantics if TLS is enabled but the private key is missing or invalid. Previous versions would only check that once connection attempts occurred. The second patch, a critical fix, enforces mutual TLS (mTLS) by default, so both parties must be keyed using the same certificate. As a workaround, avoid exposing the application publicly to avoid the vulnerability having a practical impact on one's application. However, upgrading to version 1.5.52 is still recommended by the maintainers. |
| ssh in OpenSSH before 10.1 allows the '\0' character in an ssh:// URI, potentially leading to code execution when a ProxyCommand is used. |
| The HTMLSectionSplitter class in langchain-text-splitters version 0.3.8 is vulnerable to XML External Entity (XXE) attacks due to unsafe XSLT parsing. This vulnerability arises because the class allows the use of arbitrary XSLT stylesheets, which are parsed using lxml.etree.parse() and lxml.etree.XSLT() without any hardening measures. In lxml versions up to 4.9.x, external entities are resolved by default, allowing attackers to read arbitrary local files or perform outbound HTTP(S) fetches. In lxml versions 5.0 and above, while entity expansion is disabled, the XSLT document() function can still read any URI unless XSLTAccessControl is applied. This vulnerability allows remote attackers to gain read-only access to any file the LangChain process can reach, including sensitive files such as SSH keys, environment files, source code, or cloud metadata. No authentication, special privileges, or user interaction are required, and the issue is exploitable in default deployments that enable custom XSLT. |
| The Featured Image from URL (FIFU) plugin for WordPress is vulnerable to Stored Cross-Site Scripting via a post's Featured Image custom fields in all versions up to, and including, 5.2.7 due to insufficient input sanitization and output escaping. This makes it possible for authenticated attackers, with Contributor-level access and above, to inject arbitrary web scripts in pages that will execute whenever a user accesses an injected page. NOTE: This vulnerability was partially fixed in version 5.2.2. |
| Clash Verge Rev thru 2.2.3 forces the installation of system services(clash-verge-service) by default and exposes key functions through the unauthorized HTTP API `/start_clash`, allowing local users to submit arbitrary bin_path parameters and pass them directly to the service process for execution, resulting in local privilege escalation. |
| In the Linux kernel, the following vulnerability has been resolved:
drm/radeon: Fix PCI device refcount leak in radeon_atrm_get_bios()
As comment of pci_get_class() says, it returns a pci_device with its
refcount increased and decreased the refcount for the input parameter
@from if it is not NULL.
If we break the loop in radeon_atrm_get_bios() with 'pdev' not NULL, we
need to call pci_dev_put() to decrease the refcount. Add the missing
pci_dev_put() to avoid refcount leak. |
