Variola is a modular Windows malware implant linked to a ClickFix intrusion, designed for persistent host identity and secure command-and-control communications. It supports dynamic loading of modules for credential theft, browser manipulation, and session hijacking, with capabilities typical of banking trojans such as Chromium browser injection and VNC-style remote access. The implant uses strong cryptographic methods for communication security and shows active development with multiple builds found in a public GitHub repository. No patch or remediation information is provided, and no known exploits in the wild have been reported.
Wheels CLI versions prior to 4.1.1 do not verify that the local development server responding to its HTTP requests belongs to the current project. This allows a malicious local process listening on a targeted port to receive sensitive information such as the project's reload password and code submitted via wheels reload or wheels console commands. Other commands may unintentionally affect different local applications by performing state-changing operations against their servers and databases. The vulnerability requires the developer to run an affected command while a malicious process listens on the targeted local port. This issue is fixed in version 4.1.1.
Pipecat's development runner exposes unauthenticated endpoints that can create bot sessions and paid Daily rooms if the runner is deliberately exposed to untrusted networks. The runner is designed only for local development and binds to localhost by default. Exploitation requires misconfiguration to expose it publicly. The impact includes potential resource consumption and exposure of room URLs and join tokens, but does not inherently grant access to external systems or sensitive data. Pipecat 1.6.0 clarifies this boundary and advises against using the development runner in production or exposing it to untrusted traffic.
Koha Community's Koha software contains an OS command injection vulnerability in its Task Scheduler deletion functionality. Authenticated staff users with schedule task permissions can exploit this by injecting shell metacharacters in the jobid parameter, leading to arbitrary command execution as the application user. This vulnerability is fixed in Koha versions 24.11.20, 25.05.15, 25.11.09, 26.05.04, and the 26.11 development branch.
Multiple vulnerabilities have been identified in various Microsoft development tools and frameworks, including .NET, ASP.NET, Visual Studio, Visual Studio Code, SysInternals Suite, and PowerShell. These tools are widely used for application development, system management, and automation. The specific details of the vulnerabilities, their impact, and remediation status are not provided in the available data.
Adobe ColdFusion 2025 contains a critical path traversal vulnerability (CVE-2026-90530) in its Remote Development Services (RDS) feature. This flaw allows an attacker to read arbitrary files on the file system outside the intended restricted directories without requiring user interaction. The vulnerability affects ColdFusion versions up to 2025.0.13 and 2023.0.25. RDS is intended only for development and is disabled when ColdFusion is set to Production Secure Profile. Disabling RDS or running ColdFusion in Production Secure Profile mitigates the risk.
Vitess contains a reflected cross-site scripting (XSS) vulnerability in the twopczHandler of the vttablet component. The vulnerability arises because the Action and dtid form values are reflected into the HTML response without proper HTML escaping. This flaw allows an attacker to craft URLs that execute JavaScript in the operator's browser with same-origin access to sensitive debug endpoints. The issue affects all released versions up to the current latest releases and was fixed only on the main development branch as of June 2026. No official CVE or advisory existed at the time of reporting.
FearlessCMS is a flat-file content management system centered on simplicity and open source. In 0.1.2b and earlier, an authenticated administrator can set an arbitrary Git repository URL through admin/updater-handler.php and trigger perform_update, which passes the URL to update.sh. The updater clones the repository and replaces the live plugins/, admin/, and includes/ directories without verifying the repository's identity, signatures, or contents. Because router.php permits existing PHP files under plugins to be executed by the PHP development server, attacker-supplied update content can execute arbitrary PHP with the web server process's privileges and persist as a backdoor. This issue is fixed in version 0.1.3b.
This content describes the development of a micro AI-powered Security Operations Center (SOC) for a small SaaS project. The SOC integrates Wazuh for log collection and alerting, ClickHouse for event storage and fast querying, and custom Python connectors for cloud API polling. It aims to monitor all infrastructure components, detect incidents, and enable automated or semi-automated response. The architecture is designed to run on a single VM and leverages AI for investigation and response tasks. No specific vulnerability or exploit is described.
### Summary
The Pydantic AI development web chat UI (`Agent.to_web()`, `clai web`) did not check the content type of requests to its chat endpoint. This allowed a website visited by a developer to submit a request to a chat UI running on that developer's machine, causing the served agent to run and to execute its tools with the privileges and credentials of the local process.
Binding the web UI to localhost — the default — does not prevent this, because a page open in the developer's browser can reach the loopback address.
### Impact
Applications and developers serving an agent through `Agent.to_web()` or `clai web`. The consequences depend on the tools the served agent exposes, and can include data disclosure as well as unwanted tool side effects. Tools marked `requires_approval=True` were not protected either, because the endpoint trusts approval decisions relayed by the client.
### Mitigation
Upgrade to a patched version. The chat endpoint now requires `Content-Type: application/json` and rejects other requests before the request body is parsed and before the agent runs. The bundled chat UI already sends this header; scripts and other non-browser clients that call the endpoint directly may need to be updated to send it.
If you cannot upgrade, don't run the web UI while browsing untrusted sites, stop it when you aren't using it, and don't serve an agent with side-effecting tools through it.
### Credits
- Thai Son Dinh from VinSOC Labs (R&D)
The TensorLake npm SDK version 0.5.144 was compromised in a supply chain attack involving ChainDrop/Shai-Hulud malware. The malicious package, downloaded approximately 12,000 times weekly, contains obfuscated code that steals credentials from various sources such as npm tokens, GitHub tokens, AWS credentials, HashiCorp Vault, Kubernetes configs, SSH keys, and AI development tools. It executes via a preinstall hook, establishes persistence through scheduled tasks, and supports remote code execution. The attack uses an Ethereum contract for command-and-control and includes a dead-man switch that triggers destructive actions if stolen GitHub tokens are revoked. The malware can self-propagate by republishing compromised versions to other npm packages linked to the victim's publishing identity, spreading across the npm ecosystem.
CVE-2026-76285 is a critical vulnerability in Splunk Enterprise caused by improper adherence to coding standards, categorized under CWE-710. It affects multiple versions of Splunk Enterprise prior to specific patch releases. The vulnerability has a high CVSS score of 9.8, indicating it can be exploited remotely without authentication and results in high impact on confidentiality, integrity, and availability. Cisco addressed this issue through a software hardening release as part of an internal security review.
Rockstar Games has suffered multiple confirmed security compromises from 2018 through 2026, involving diverse attack vectors rather than zero-day exploits. These include MFA fatigue attacks, exploitation of hardcoded credentials in collaboration tools, stolen OAuth tokens from third-party SaaS vendors, and reverse-engineered P2P network vulnerabilities leading to remote code execution. The most severe incident in August 2026 involved exfiltration of a playable GTA VI development build and extensive gameplay data due to failures in internal segmentation and data loss prevention. The attacks highlight systemic weaknesses in identity management, token security, and internal network controls.
A moderate information disclosure vulnerability (CVE-2026-58649) exists in the .NET SDK used in Red Hat Hardened Images. The vulnerability arises from insufficient origin validation in the dotnet watch BrowserRefreshServer, allowing a remote unauthenticated attacker to connect to a development WebSocket and retrieve compiled intermediate language assemblies and debug files. This could expose confidential information from the development environment. Red Hat rates this issue as medium severity with a CVSS 3.1 score of 6.5. A fix is currently being pursued but not yet available, and no effective mitigation meeting Red Hat's criteria is currently provided.
Quasar Framework is a framework for building high-performance Vue.js user interfaces. Prior to @quasar/render-ssr-error 2.2.4 and @quasar/app-vite 3.3.0, renderSSRError() in utils/render-ssr-error/src/index.js used diagnostic data from utils/render-ssr-error/src/env.js to serialize process.env, request headers, and cookies into the HTTP page returned by serve.devError(), while the development server listened on all interfaces by default. Any network-adjacent client that reaches an SSR or SSG render failure through this development-only error path can obtain shell environment secrets. The renderer escaped only one exact lowercase script closing-tag spelling, so case variants and valid closing-tag delimiter variants in reflected diagnostic data could terminate the script element and inject markup; executing the injected code in a developer browser additionally requires the payload to accompany that developer's request. This issue is fixed in @quasar/render-ssr-error 2.2.4 and @quasar/app-vite 3.3.0.
Quasar Framework is a framework for building high-performance Vue.js user interfaces. Prior to @quasar/ssl-certificate 2.1.0, @quasar/cli 5.0.4, and @quasar/app-vite 3.3.0, the @quasar/ssl-certificate utility cached a combined private key and certificate PEM without explicitly applying owner-only filesystem permissions. Another local user able to read the cache can copy the key and impersonate a development TLS endpoint in an environment that trusts the certificate. The generated certificate was also CA-capable, carried unnecessarily broad key usages, and encoded the IPv6 loopback address as a DNS subject alternative name. This issue is fixed in @quasar/ssl-certificate 2.1.0, @quasar/cli 5.0.4, and @quasar/app-vite 3.3.0.
This content discusses how frontier AI models are transforming vulnerability management, particularly for Microsoft software. AI accelerates vulnerability discovery and patch development, resulting in a significant increase in identified vulnerabilities, especially for on-premises Microsoft products. Microsoft mitigates most cloud vulnerabilities without customer action, but on-premises customers should expect more frequent and numerous patches. CISOs face challenges balancing rapid patching with accuracy at unprecedented scale. The post recommends increased patching resources, faster patch deployment for critical systems, use of AI-powered scanning harnesses, and defense-in-depth strategies. Microsoft also collaborates with industry peers to address open-source vulnerabilities and promotes Secure by Design and Secure by Default principles to improve security posture.
AI tools enable non-technical employees, known as 'citizen coders,' to rapidly build workplace applications using natural language prompts. While these AI-generated apps can enhance productivity and reduce developer workload, they often bypass IT and security oversight, creating significant security and compliance risks. These unsanctioned applications may contain critical vulnerabilities, misconfigurations, weak data protection, excessive permissions, and lack proper maintenance such as patching or monitoring. The volume of such apps can overwhelm IT and security teams, complicating governance. Tenable recommends a structured, multi-tiered governance framework combining peer leadership and mandatory security training to manage these risks effectively, rather than imposing outright bans which may drive development underground.