CVE Feed

SGLang's multimodal generation runtime is vulnerable to unauthenticated arbitrary code execution because the disaggregated-diffusion orchestrator's DiffusionServer binds an unauthenticated ZeroMQ ROUTER socket to a network interface and passes the final frame of received multipart messages directly to pickle.loads() before any validation occurs.

NumberInput.looksLikeValidNumber() in FasterXML jackson-core pre-validates "stringified numbers" with two regular expressions: PATTERN_FLOAT ([+-]?[0-9]*[\.]?[0-9]+([eE][+-]?[0-9]+)?), present since 2.17.0, and PATTERN_FLOAT_TRAILING_DOT, added in 2.17.2. PATTERN_FLOAT places adjacent quantifiers over the same character class -- an optional [0-9]* run, an optional dot, then a required [0-9]+ run -- so input that ultimately fails to match forces Java's backtracking engine to retry every possible split point of the digit run.  Matching cost therefore grows with the square of the input length.  An attacker who can supply JSON that an application deserializes into a numeric target type reaches this method through jackson-databind's default String-to-number coercion (StdDeserializer and NumberDeserializers for BigDecimal, BigInteger, Double and Float).  Because StreamReadConstraints.maxStringLength defaults to 20,000,000 characters, no constraint bounds the input before it reaches the regex.  Testing by the reporter confirmed O(n^2) growth across five consecutive input-size doublings, with a single 160,000-character string consuming roughly 74 seconds in one call; a small number of concurrent requests of ordinary body size can therefore exhaust a server's request-handling thread pool.  The affected method does not exist before 2.17.0, so 2.16.x and earlier releases are not affected.  The fix replaces both regular expressions with a hand-rolled single-pass scan.

A improper restriction of rendered ui layers or frames vulnerability in Fortinet FortiPAM Chrome Extension 8.0 all versions, FortiPAM Chrome Extension 7.4 all versions may allow attacker to information disclosure via remote unauthenticated attack

A reflected cross-site scripting vulnerability exists in x-ui 0.3.2. The management interface reflects the raw request URI into a client-side template binding expression used for sidebar menu highlighting. Server-side HTML entity escaping is ineffective in this context: the browser decodes the entities before the client-side framework evaluates the content as a JavaScript expression. A logged-in panel user who visits a crafted URL allows arbitrary script execution in the same-origin context of the management page, enabling data theft and unauthorized actions through the victim's session.

A horizontal privilege escalation vulnerability exists in x-ui 0.3.2. An authenticated user can modify the inbound proxy configurations of other users, including remark, port, protocol, settings, enabled state, expiry time and traffic quota, by submitting a request referencing the target resource identifier. The update path fails to verify that the target resource belongs to the requesting session user, allowing unauthorized cross-user modification of data.

webpy web.py 0.76 is vulnerable to Insufficient Session Expiration. The application's session management relies on periodic cleanup to expire sessions instead of checking the last-access time when a session is loaded. As a result, an expired session whose record has not yet been cleaned up can still be replayed and used, allowing an attacker holding a previously valid session cookie to continue accessing protected resources after the configured idle timeout.

NVIDIA NeMo contains a vulnerability in the TabularTokenizer class where it deserializes an untrusted, attacker-controlled .pkl file via pickle.load() without validation. A successful exploit of this vulnerability may lead to code execution, data tampering, denial of service, and information disclosure.

NVIDIA NeMo contains a vulnerability in its dataset-loading workflow where a maliciously crafted model_config.yaml can inject unsafe parameters. A successful exploit of this vulnerability may lead to code execution, data tampering, denial of service, and information disclosure.

NVIDIA Infrastructure Controller for Linux contains a vulnerability where an attacker could cause OS command injection. A successful exploit of this vulnerability might lead to code execution, data tampering, denial of service, and information disclosure.

NVIDIA Infrastructure Controller for Linux contains a vulnerability where an attacker could cause improper certificate validation. A successful exploit of this vulnerability might lead to information disclosure, data tampering, and denial of service.

NVIDIA Infrastructure Controller for Linux contains a vulnerability where an attacker could cause SQL injection. A successful exploit of this vulnerability might lead to code execution, data tampering, denial of service, and information disclosure.

NVIDIA Infrastructure Controller for Linux contains a vulnerability where an attacker could cause exposure of sensitive system information due to uncleared debug information. A successful exploit of this vulnerability might lead to information disclosure.

NVIDIA Infrastructure Controller for Linux contains a vulnerability where an attacker could cause improper enforcement of a behavioral workflow. A successful exploit of this vulnerability might lead to data tampering, denial of service, and information disclosure.

NVIDIA Infrastructure Controller for Linux contains a vulnerability where an attacker could cause an XML injection. A successful exploit of this vulnerability might lead to data tampering and denial of service.

NVIDIA Infrastructure Controller for Linux contains a vulnerability where an attacker could cause an improper authentication issue. A successful exploit of this vulnerability might lead to escalation of privileges, information disclosure, and data tampering.

NVIDIA Infrastructure Controller for Linux contains a vulnerability where an attacker could cause improper certificate validation. A successful exploit of this vulnerability might lead to information disclosure, data tampering, and denial of service.

NVIDIA Infrastructure Controller for Linux contains a vulnerability where an attacker could cause use of a hard-coded password. A successful exploit of this vulnerability might lead to data tampering, denial of service, and information disclosure.

NVIDIA Infrastructure Controller for Linux contains a vulnerability where an attacker may cause uncontrolled resource consumption. A successful exploit of this vulnerability may lead to denial of service.

NVIDIA Infrastructure Controller for Linux contains a vulnerability where an attacker could cause missing authentication for a critical function. A successful exploit of this vulnerability might lead to data tampering, denial of service, and information disclosure.

NVIDIA Infrastructure Controller for Linux contains a vulnerability where an attacker could cause use of hard-coded credentials. A successful exploit of this vulnerability might lead to escalation of privileges, data tampering, denial of service, and information disclosure.

NVIDIA Infrastructure Controller for Linux contains a vulnerability where an attacker could cause uncontrolled resource consumption. A successful exploit of this vulnerability might lead to denial of service.

NVIDIA NeMo Speech for all platforms contains a vulnerability where malicious input created by an attacker could cause a code injection. A successful exploit of this vulnerability might lead to code execution, information disclosure, and data tampering.

NVIDIA NeMo Speech for all platforms contains a vulnerability in the speech data explorer component, where malicious data created by an attacker could cause remote code execution. A successful exploit of this vulnerability might lead to code execution, escalation of privileges, information disclosure, and data tampering.

NVIDIA NeMo Speech for all platforms contains a vulnerability where malicious data created by an attacker could cause remote code execution. A successful exploit of this vulnerability might lead to code execution, information disclosure, and data tampering.

A potential security vulnerability has been identified in the HP Support Assistant for versions prior to 9.55.10.0. The vulnerability could potentially allow a local attacker to escalate privileges due to insufficient access controls.

MISP contains a stored cross-site scripting (XSS) vulnerability in the admin email composition screen. The MISP.org organization name setting was interpolated directly into a JavaScript string literal using an unescaped PHP echo: var org = "<?php echo $org;?>";. Because the value was placed inside a double-quoted JavaScript string without any encoding, an organization name containing a double-quote character (or a backslash) could terminate the string literal and inject arbitrary JavaScript into the page. The injected script would execute in the context of any authenticated user who subsequently loads the admin email page, potentially allowing session hijacking, data exfiltration, or privileged actions performed on behalf of the victim. Exploitation requires the ability to set or modify the MISP.org organization name and a second authenticated user visiting the affected admin email view. The vulnerability is a classic instance of insufficient output encoding in a JavaScript context.

MISP's RequestHandlerComponent automatically decodes XML request bodies on all write requests. The underlying Xml::build() library contains a logic error in its readFile guard condition (readFile && http || https), where PHP operator precedence causes the https branch to bypass the readFile check entirely. As a result, a request body containing a bare HTTPS URL is treated as a locator and fetched by the server. Because the cspReport endpoint accepts XML content type and is unauthenticated by design, an unauthenticated remote attacker can send a POST request with a crafted XML body to cause the MISP server to issue an outbound HTTPS request to an arbitrary target reachable from the server. The attacker does not receive the response body, but can use the vulnerability to probe internal services, trigger actions on internal HTTPS endpoints, or perform timing-based reconnaissance of the server's network environment.  The vulnerability affects any MISP deployment where the cspReport endpoint is reachable and the server has network access to internal or external HTTPS services.

D-Link DAP-1360 firmware version 6.14 and earlier contains an unauthenticated remote code execution vulnerability that allows remote attackers to execute arbitrary commands as root by sending crafted requests to the device's web management interface without valid credentials. Attackers can fully compromise the device to persistently modify its configuration and use it as a pivot point into the local network.

In MISP, the queryEnrichment method in EventsController.php accepted a module name parameter and iterated over the list of enabled modules to find a match. If the specified module was not present in the enabled modules list, the code silently continued processing using default parameters (format set to 'simplified' and no module-specific configuration applied) rather than rejecting the request. This allowed an authenticated user to trigger enrichment or analysis processing through a module that was not explicitly enabled or available on the instance, bypassing the intended module availability controls. The lack of validation meant that module names outside the enabled set were not rejected, potentially exposing functionality or data processing paths that the administrator had not authorized for use.

In MISP, the CollectionsController add() method enforced the sharing-group usability authorization check and element capture only when the HTTP request method was POST. However, the underlying CRUDComponent::add() method persists data on both POST and PUT requests. As a result, an authenticated user could issue a PUT request to the collections/add endpoint, bypassing the sharing-group authorization check entirely. This allowed the creation of a collection with distribution=4 (shared with a specific sharing group) targeting a sharing group that the user was not authorized to use. The bypassed check would normally verify that the current user has access to the specified sharing group before allowing the collection to be associated with it. The security impact is that an authenticated user can place collection data into a sharing group context they do not have legitimate access to, potentially exposing that data to all members of the unauthorized sharing group and violating the intended access-control boundaries of the MISP instance. The vulnerability requires an authenticated session and knowledge of a valid sharing group identifier; no user interaction or special tooling is needed beyond sending a PUT request instead of a POST request to the same endpoint.

The MISP installer scripts (for Debian 12, Debian 13, Ubuntu 24.04, and RHEL 9.4) create a log file at /var/log/misp_install.log and a named pipe (FIFO) at /var/log/misp_install.log.pipe to capture all installer output. The log captures highly sensitive data including the generated admin password, database passwords, GPG passphrase, and supervisor password. However, the log file was created by the tee command under the system default umask, resulting in world-readable permissions (typically 0644). Similarly, the FIFO was created with mkfifo without an explicit mode, also inheriting the default umask and remaining world-readable during the window before it was unlinked.  Any local unprivileged user on the system could read the log file or the FIFO to obtain these credentials.  Additionally, the log file was not removed before creation, meaning a pre-existing symlink in /var/log could have been used to redirect the write to an arbitrary location. Version affected: <2.5.47

Mattermost versions 11.9.x <= 11.9.1, 11.8.x <= 11.8.5, 11.7.x <= 11.7.10, 11.10.x <= 11.10.1 fail to limit the length of the post ID array accepted by the bulk reactions endpoint which allows an authenticated user to cause excessive database load via a crafted request to {{POST /api/v4/posts/ids/reactions}}.. Mattermost Advisory ID: MMSA-2026-00771

MISP contains a reflected cross-site scripting (XSS) vulnerability in the event REST search export confirmation form. The view template app/View/Events/ajax/eventRestSearchExportConfirmationForm.ctp renders a URL-supplied event ID list into a single-quoted JavaScript string literal using PHP's json_encode() without any hex-encoding flags. By default, json_encode() escapes double quotes and backslashes but does not escape single quotes. Because the JavaScript string is delimited by single quotes, an attacker can inject a single-quote character to terminate the string literal and execute arbitrary JavaScript in the victim's browser session. The vulnerability affects the Default and UiBeta themes, both of which render this view. The Overmind theme's own copy of the form already passed the value through an escaped data attribute and was not affected by this specific sink. Preconditions: the victim must be an authenticated MISP user and must actively open or navigate to the attacker-crafted URL. The attacker does not require an account or any prior access to the MISP instance. Security impact: successful exploitation allows execution of arbitrary JavaScript in the context of the MISP web application, potentially leading to session hijacking, unauthorized actions performed on behalf of the victim, exfiltration of sensitive data visible in the session, or further client-side attacks. The vulnerable component's own confidentiality, integrity, and availability are not directly compromised; the impact is on the underlying user session.

A flaw has been found in JosephChuks php-file-manager-with-code-editor up to 3.0. This issue affects the function move_uploaded_file of the file filemanager.php. Executing a manipulation of the argument files can lead to unrestricted upload. The attack can be executed remotely. The vendor was contacted early about this disclosure but did not respond in any way.

A vulnerability was identified in sfturing hosp_order up to 627f426331da8086ce8fff2017d65b1ddef384f8. Affected is an unknown function of the file HospitalController.java of the component Public Search Handlers. The manipulation of the argument Search leads to cross site scripting. The attack can be initiated remotely. The exploit is publicly available and might be used. The project was informed of the problem early through an issue report but has not responded yet.

MarketKing plugin for WordPress before 2.1.72 contains a missing authorization vulnerability in the marketking_admin_vendors_ajax AJAX action that allows authenticated attackers with subscriber-level access or higher to retrieve the complete vendor directory by sending a crafted AJAX request. Attackers can exploit the absence of capability checks in the vendor management action to retrieve internal user IDs, usernames, and email addresses of all registered vendors, exposing personally identifiable information to any logged-in user regardless of role.

MarketKing plugin for WordPress before 2.1.72 contains a missing authorization vulnerability in the marketking_duplicate_product AJAX action that allows authenticated attackers with subscriber-level access or higher to duplicate any vendor's product by supplying an arbitrary product ID. Attackers can bypass ownership verification to copy any vendor's product listings, including private product metadata, and assign the duplicated copy to their own vendor account without the victim's knowledge or consent.

MarketKing plugin for WordPress before 2.1.72 contains a missing authorization vulnerability in the marketking_send_refund AJAX action that allows authenticated attackers with subscriber-level access or higher to create refund requests against any order by supplying an arbitrary order ID. Attackers can submit crafted AJAX requests targeting any order ID to create fraudulent refund requests against orders they did not place, enabling marketplace disruption and unauthorized interference with other users' orders.

Improper neutralization of special elements used in an SQL command ('SQL injection') vulnerability in Karel Electronic Industry and Trade Inc. KarelIPS allows Blind SQL Injection. This issue affects KarelIPS: through 22092026. NOTE: The vendor was contacted and it was learned that the product is not supported.