CVE-2025-14927 is a high-severity remote code execution vulnerability in Hugging Face Transformers version 4.57.0, specifically in the convert_config function. It arises from improper validation of user-supplied strings that are executed as Python code, enabling attackers to execute arbitrary code with the privileges of the current user. Exploitation requires user interaction, specifically converting a malicious checkpoint file. The vulnerability impacts confidentiality, integrity, and availability, with a CVSS score of 7.8. There are no known exploits in the wild yet, but the risk is significant due to the widespread use of Hugging Face Transformers in AI and ML workflows. European organizations using this library in development or production environments are at risk, especially those in countries with strong AI sectors. Mitigation involves avoiding untrusted checkpoint files, applying patches once available, and implementing strict input validation and sandboxing.
CVE-2025-14926 is a high-severity code injection vulnerability in Hugging Face Transformers version 4.57.0, specifically in the convert_config function. It allows remote attackers to execute arbitrary Python code by supplying a malicious checkpoint configuration. Exploitation requires user interaction, as the target must convert a crafted checkpoint file. The vulnerability arises from improper validation of user-supplied strings before execution, enabling code execution with the privileges of the current user. Although no known exploits are currently in the wild, the impact on confidentiality, integrity, and availability is high. This vulnerability poses a significant risk to organizations using Hugging Face Transformers for machine learning model management or deployment. European organizations relying on this library, especially in AI research and development sectors, should prioritize patching or mitigating this issue. Countries with strong AI and tech industries, such as Germany, France, the UK, and the Netherlands, are likely to be most affected.
Dive is an open-source MCP Host Desktop Application that enables integration with function-calling LLMs. A critical Stored Cross-Site Scripting (XSS) vulnerability exists in versions prior to 0.11.1 in the Mermaid diagram rendering component. The application allows the execution of arbitrary JavaScript via `javascript:`. An attacker can exploit this to inject a malicious Model Context Protocol (MCP) server configuration, leading to Remote Code Execution (RCE) on the victim's machine when the node is clicked. Version 0.11.1 fixes the issue.
AVideo versions 14.3.1 prior to 20.1 contain an unauthenticated remote code execution vulnerability caused by predictable generation of an installation salt using PHP uniqid(). The installation timestamp is exposed via a public endpoint, and a derived hash identifier is accessible through unauthenticated API responses, allowing attackers to brute-force the remaining entropy. The recovered salt can then be used to encrypt a malicious payload supplied to a notification API endpoint that evaluates attacker-controlled input, resulting in arbitrary code execution as the web server user.
The Ocean Modal Window WordPress plugin before 2.3.3 is vulnerable to Remote Code Execution via the modal display logic. These modals can be displayed under user-controlled conditions that Editors and Administrators can set (edit_pages capability). The conditions are then executed as part of an eval statement executed on every site page. This leads to remote code execution.
ChurchCRM is an open-source church management system. Prior to version 5.21.0, a pre-authentication remote code execution vulnerability in ChurchCRM's setup wizard allows unauthenticated attackers to inject arbitrary PHP code during the initial installation process, leading to complete server compromise. The vulnerability exists in `setup/routes/setup.php` where user input from the setup form is directly concatenated into a PHP configuration template without any validation or sanitization. Any parameter in the setup form can be used to inject PHP code that gets written to `Include/Config.php`, which is then executed on every page load. This is more severe than typical authenticated RCE vulnerabilities because it requires no credentials and affects the installation process that administrators must complete. Version 5.21.0 patches the issue.
DeepChat is an open-source artificial intelligence agent platform that unifies models, tools, and agents. Prior to version 0.5.3, a security vulnerability exists in the Mermaid diagram rendering component that allows arbitrary JavaScript execution. Due to the exposure of the Electron IPC renderer to the DOM, this Cross-Site Scripting (XSS) flaw escalates to full Remote Code Execution (RCE), allowing an attacker to execute arbitrary system commands. Two concurrent issues, unsafe Mermaid configuration and an exposed IPC interface, cause this issue. Version 0.5.3 contains a patch.
Fickling is a Python pickling decompiler and static analyzer. Versions prior to 0.1.6 had a bypass caused by `pty` missing from the block list of unsafe module imports. This led to unsafe pickles based on `pty.spawn()` being incorrectly flagged as `LIKELY_SAFE`, and was fixed in version 0.1.6. This impacted any user or system that used Fickling to vet pickle files for security issues.
The The Shortcode Ajax plugin for WordPress is vulnerable to arbitrary shortcode execution in all versions up to, and including, 1.0. This is due to the software allowing users to execute an action that does not properly validate a value before running do_shortcode. This makes it possible for unauthenticated attackers to execute arbitrary shortcodes.
CVE-2025-67750 is a high-severity code injection vulnerability in lightning-flow-scanner versions prior to 6.10.6. It arises from improper control of code generation (CWE-94) where maliciously crafted Salesforce flow metadata files can trigger arbitrary JavaScript execution during scanning. The vulnerability exploits the APIVersion rule's use of new Function() to evaluate expression strings, allowing attackers to inject malicious code. This can compromise developer machines, CI/CD runners, and editor environments using the vulnerable tool. No authentication or user interaction is required to trigger the flaw, and it affects local environments where the scanner is run. Although no known exploits are reported in the wild, the impact on confidentiality, integrity, and availability is high. The issue is fixed in version 6.10.
Parse Server is an open source backend that can be deployed to any infrastructure that runs Node.js. In versions prior to 8.6.0-alpha.2, a GitHub CI workflow is triggered in a way that grants the GitHub Actions workflow elevated permissions, giving it access to GitHub secrets and write permissions which are defined in the workflow. Code from a fork or lifecycle scripts is potentially included. Only the repository's CI/CD infrastructure is affected, including any public GitHub forks with GitHub Actions enabled. This issue is fixed version 8.6.0-alpha.2 and commits 6b9f896 and e3d27fe.
The WPMasterToolKit plugin for WordPress is vulnerable to PHP Code Injection in all versions up to, and including, 2.13.0. This is due to the plugin allowing Author-level users to create and execute arbitrary PHP code through the Code Snippets feature without proper capability checks. This makes it possible for authenticated attackers, with Contributor-level access and above, to execute arbitrary PHP code on the server, leading to remote code execution, privilege escalation, and complete site compromise.
Neuron is a PHP framework for creating and orchestrating AI Agents. Versions 2.8.11 and below use MySQLSelectTool, which is vulnerable to Read-Only Bypass. MySQLSelectTool is intended to be a read-only SQL tool (e.g., for LLM agent querying, however, validation based on the first keyword (e.g., SELECT) and a forbidden-keyword list does not block file-writing constructs such as INTO OUTFILE / INTO DUMPFILE. As a result, an attacker who can influence the tool input (e.g., via prompt injection through a public agent endpoint) may write arbitrary files to the DB server if the MySQL/MariaDB account has the FILE privilege and server configuration permits writes to a useful location (e.g., a web-accessible directory). This issue is fixed in version 2.8.12.
@vitejs/plugin-rs provides React Server Components (RSC) support for Vite. Versions 0.5.5 and below are vulnerable to arbitrary remote code execution on the development server through unsafe dynamic imports in server function APIs (loadServerAction, decodeReply, decodeAction) when integrated into RSC applications that expose server function endpoints. Attackers with network access to the development server can read/modify files, exfiltrate sensitive data (source code, environment variables, credentials), or pivot to other internal services. While this affects development servers only, the risk increases when using vite --host to expose the server on all network interfaces. This issue is fixed in version 0.5.6.
Elysia is a Typescript framework for request validation, type inference, OpenAPI documentation and client-server communication. Versions 1.4.17 and below are subject to arbitrary code execution from cookie config. When dynamic cookies are enabled (e.g.there an existing cookie schema), the cookie config is injected into the compiled route without first being sanitised. Availability of this exploit is generally low, but when combined with GHSA-hxj9-33pp-j2cc, it allows for a full RCE chain. An attack requires write access to either the Elysia app's source code (in which case the vulnerability is meaningless) or write access to the cookie config (perhaps where it is assumed to be provisioned by the environment). This issue is fixed in version 1.4.18.
The Paid Membership Plugin, Ecommerce, User Registration Form, Login Form, User Profile & Restrict Content – ProfilePress plugin for WordPress is vulnerable to arbitrary shortcode execution in all versions up to, and including, 4.16.7 due to insufficient input sanitization on the `type` parameter in the form preview functionality. This makes it possible for authenticated attackers, with Subscriber-level access and above, to execute arbitrary shortcodes via the `pp_preview_form` endpoint.
Improper Control of Generation of Code ('Code Injection') vulnerability in Liquid Web / StellarWP GiveWP give allows Code Injection.This issue affects GiveWP: from n/a through 4.13.1.
JIT miscompilation in the JavaScript Engine: JIT component. This vulnerability was fixed in Firefox 146, Firefox ESR 115.31, Firefox ESR 140.6, Thunderbird 146, and Thunderbird 140.6.
