A command injection vulnerability exists in the documentserver-flush-cache.sh maintenance helper script in ONLYOFFICE Docs (Document Server) version 9.4.0-129. A local attacker with ds user privileges can exploit this flaw by supplying a crafted --hash argument that is improperly validated and interpolated into a GNU sed expression. This allows execution of arbitrary commands as root due to the sed e flag and the sudoers policy permitting ds to run the helper as root without a password and with unrestricted arguments.
An out-of-bounds read vulnerability exists in VideoLAN VLC 3.0.23. By using a specially crafted NUV file, an attacker could exploit this vulnerability to perform an out-of-bounds read, potentially leading to the disclosure of sensitive information.
NodePress versions up to 8.5.1 have an authentication bypass vulnerability due to a hardcoded default password 'root'. Unauthenticated attackers can exploit this by sending a base64-encoded 'root' password to the /admin/login endpoint when the NP_ADMIN_DEFAULT_PASSWORD environment variable is unset, allowing them to obtain administrator tokens and gain full control of the site. This vulnerability is fixed in version 8.5.2.
NodePress versions up to 8.5.1 have a rate limit bypass vulnerability due to trustProxy being enabled. This allows unauthenticated attackers to spoof the X-Forwarded-For header to evade throttling mechanisms. Attackers can exploit this to brute-force administrator passwords via the /admin/login endpoint and bypass comment IP blocklists. The vulnerability is fixed in version 8.5.2.
Rowy versions up to 3.0.0 have a vulnerability in their default Firebase Storage rules that incorrectly assign permissions. This flaw allows any user with a role to gain read and write access to the entire storage bucket. Consequently, authenticated users who should have limited or read-only access can exploit the Firebase SDK to list, download, overwrite, or delete any files in the bucket. The issue is fixed in version 3.0.1.
Prime CMS versions prior to 0.4.0-beta.0 contain an OS command injection vulnerability in the updateSystem GraphQL mutation. Authenticated administrators can exploit this flaw by injecting shell metacharacters into the package version string, which is executed by the server process via child_process.exec() during yarn add or npm install commands. This allows execution of arbitrary shell commands with the privileges of the server process. The vulnerability is fixed in version 0.4.0-beta.0.
Rocboss paopao-ce versions 0.2.0 through 0.6.0-alpha.3 have a path traversal vulnerability in the LocalOSS DeleteObject function. Authenticated users can exploit this flaw by submitting crafted media values containing dot-dot-slash sequences to endpoints such as POST /v1/user/avatar or POST /v1/post/comment. This allows deletion of arbitrary server-writable files, including configuration and data files.
Dinky versions up to 1.2.5 have a missing authorization vulnerability in the ClusterConfigurationController. This flaw allows authenticated users without proper cluster configuration permissions to access and list Flink cluster configurations. Attackers can exploit this by querying specific API endpoints to retrieve full deployment configuration details for every cluster within their tenant.
LimeSurvey versions before 6.17.16+260814 and 7.x before 7.0.10+260813 have an unauthenticated SQL injection vulnerability in the public statistics page. This flaw allows attackers to submit specially crafted POST parameter names containing backticks to exploit the buildSelects() function. Exploitation can lead to reading administrator password hashes and survey responses on MySQL or MariaDB databases.
An improper privilege management vulnerability exists in an LDAP extension used with OpenText eDirectory. This flaw could cause a technical or proxy user to receive write permission to the ACL instead of just permission to modify a specific LDAP attribute. An attacker with the proxy user's credentials could exploit this to modify access rights within the affected container, potentially leading to unauthorized privilege escalation. The vulnerability is addressed in OpenText eDirectory version 25.2.
CVE-2026-127044 is an Insecure Direct Object Reference (IDOR) vulnerability in the STEL Order Tempo module. The vulnerability arises because the server retrieves employee data based on the empleadoId parameter without verifying the requesting user's permissions. An authenticated user can exploit this flaw to access personal information of other employees, including first name, last name, role, job title, and vacation days.
Unrolled Secure versions before 1.18.0 have an SSLRedirect bypass vulnerability in the Secure.isSSL() function. This function incorrectly trusts the client-supplied r.URL.Scheme to determine if the transport is HTTPS. An attacker can exploit this by sending plaintext HTTP requests with an absolute-form https:// request-target, bypassing the HTTPS redirect and causing responses to be served unencrypted.
Chinese threat actors have been using the Artex agentic AI penetration testing framework to accelerate cyberattacks, including a recent breach of Barracuda. Artex automates complex multi-target operations, enabling attackers with minimal expertise to conduct rapid, large-scale intrusions. The Barracuda attack involved exploitation of exposed cloud access keys, credential theft, lateral movement, and persistence, with attackers accessing sensitive cloud accounts and leaving backdoors in Kubernetes clusters. The framework has been linked to multiple campaigns targeting government, financial, AI development, and cryptocurrency sectors globally. The FBI and security researchers warn that Artex-powered attacks leverage automation to outpace defenders and exploit common security hygiene failures such as exposed credentials and unpatched software.
@fastify/busboy versions 3.2.1 through 3.2.5 contain an off-by-one error in multipart form data parsing that can cause a file attachment part to be misinterpreted as a text field. This occurs when the multipart header terminator sequence is split exactly at the configured header size limit, dropping the final header line. A remote attacker controlling the multipart body can exploit this to bypass upload validation that only applies to file events, potentially allowing malicious payloads to be processed incorrectly. The issue is fixed in version 3.2.6.
Kunai, a security monitoring tool, contains a time-of-check to time-of-use (TOCTOU) race condition in its mount namespace (mnt ns) caching logic. When resolving a security event that references a specific process PID and an expected mount namespace, the Switcher::new function read /proc/<pid>/ns/mnt to determine the target namespace and cached the result under the expected namespace key without verifying that the resolved namespace actually matched the expected one.
An attacker with local access could exploit this by either (1) arranging for the referenced PID to be reused by a process in a different namespace (e.g., the host namespace) before Kunai resolves it, or (2) calling unshare or setns on the target process after the event is generated but before the namespace is resolved. In either case, the wrong namespace (potentially the host's own namespace) would be cached under the container's namespace key.
The security impact is that subsequent container security events would be resolved against host filesystem paths rather than container paths. This undermines the integrity of the security monitoring data: container activity could be misattributed to the host, host activity could be misattributed to a container, and file-path-based detections or alerts would reference incorrect paths. An attacker could potentially use this to obscure malicious activity within a container by making it appear as benign host activity, or to generate misleading security alerts.
The vulnerability requires local access to the system and a specific timing window between event generation and namespace resolution. No remote attack vector or user interaction is required.
Requests targeting the /mcp-rest/test/tools/list endpoint have been observed attempting to exploit two vulnerabilities, CVE-2026-42271 and CVE-2026-48710, affecting LiteLLM and the Starlette Python web framework. The root cause involves Starlette versions prior to 1.0.1 not validating the Host header, allowing attackers to manipulate access control logic and potentially achieve remote code execution without authentication. Exploitation attempts use malformed Host headers to bypass filters. This vulnerability has been known since at least July 2026, and public exploit code is available via Metasploit.
Kaspersky reported suspected exploitation of a vulnerability in Telegram's newly launched widgets feature to deliver WebRAT malware. Researchers demonstrated remote code execution (RCE) exploits shortly after the feature's release. The malware, WebRAT, is a Rust-based trojan capable of system control and stealing cryptocurrency wallet and Telegram account data. Telegram's CEO disputed the claims, stating that no one could be hacked just by receiving a message and that the attack required user interaction with a third-party service, which has since been patched server-side. Telegram also fixed an Android RCE issue and deployed the fix to all users. Researchers demonstrated a second RCE exploit on Windows, which Telegram has not publicly confirmed as fixed. Kaspersky recommends not launching widgets until a fix is released.
CVE-2026-21150 is an improper access control vulnerability in edmnativehelper prior to SMR Nov-2026 Release 1. It allows local attackers to access sensitive information. The vulnerability has a medium severity rating with a CVSS score of 5.1. No exploit in the wild or patch information is currently available.
CVE-2026-63245 is a local privilege escalation vulnerability in the Emupia2k.sys driver file of the Audigy Rx device. A local attacker with low privileges could exploit this flaw to gain higher privileges, including administrator rights. The vulnerability has a high severity rating with a CVSS score of 7.3. No affected versions or patch information are currently provided, and no known exploits are reported in the wild.
A vulnerability in the Linux kernel's FUSE subsystem could cause a kernel panic when the copy buffer is exhausted during io-uring operations. This occurs because fuse_copy_fill() triggers a BUG_ON condition if the iterator backing the copy runs out of pages. An unprivileged user can exploit this by mounting a FUSE filesystem with small payload buffers and then setting an extended attribute with a larger value, causing the kernel to panic. The issue has been resolved by changing the behavior to return an error code instead of triggering a BUG.