CWE-470

The vSphere Client (HTML5) contains a remote code execution vulnerability due to lack of input validation in the Virtual SAN Health Check plug-in which is enabled by default in vCenter Server. A malicious actor with network access to port 443 may exploit this issue to execute commands with unrestricted privileges on the underlying operating system that hosts vCenter Server.

In yiisoft/yii2 version 2.0.48, the base Component class contains a vulnerability where the `__set()` magic method does not validate that the value passed is a valid Behavior class name or configuration. This allows an attacker to instantiate arbitrary classes, passing parameters to their constructors and invoking setter methods. Depending on the installed dependencies, various types of attacks are possible, including the execution of arbitrary code, retrieval of sensitive information, and unauthorized access.

An unsafe reflection vulnerability was identified in GitHub Enterprise Server that could lead to reflection injection. This vulnerability could lead to the execution of user-controlled methods and remote code execution. To exploit this bug, an actor would need to be logged into an account on the GHES instance with the organization owner role. This vulnerability affected all versions of GitHub Enterprise Server prior to 3.12 and was fixed in versions 3.8.13, 3.9.8, 3.10.5, and 3.11.3. This vulnerability was reported via the GitHub Bug Bounty program.

Horde Groupware Webmail Edition through 5.2.22 allows a reflection injection attack through which an attacker can instantiate a driver class. This then leads to arbitrary deserialization of PHP objects.

In Progress® Telerik® UI for AJAX, versions 2011.2.712 to 2025.1.218, an unsafe reflection vulnerability exists that may lead to an unhandled exception resulting in a crash of the hosting process and denial of service.

On F5 BIG-IP 13.1.0-13.1.0.3 or 13.0.0, when authenticated administrative users execute commands in the Traffic Management User Interface (TMUI), also referred to as the BIG-IP Configuration utility, restrictions on allowed commands may not be enforced.

Use of Externally-Controlled Input to Select Classes or Code ('Unsafe Reflection') vulnerability in Sitecore Sitecore Experience Manager (XM), Sitecore Experience Platform (XP) allows Cache Poisoning.This issue affects Sitecore Experience Manager (XM): from 9.0 through 9.3, from 10.0 through 10.4; Experience Platform (XP): from 9.0 through 9.3, from 10.0 through 10.4.

Legion of the Bouncy Castle Legion of the Bouncy Castle Java Cryptography APIs 1.58 up to but not including 1.60 contains a CWE-470: Use of Externally-Controlled Input to Select Classes or Code ('Unsafe Reflection') vulnerability in XMSS/XMSS^MT private key deserialization that can result in Deserializing an XMSS/XMSS^MT private key can result in the execution of unexpected code. This attack appear to be exploitable via A handcrafted private key can include references to unexpected classes which will be picked up from the class path for the executing application. This vulnerability appears to have been fixed in 1.60 and later.

Check Point Endpoint before version E86.50 failed to protect against specific registry change which allowed to disable endpoint protection by a local administrator.

Those using java.sql.Statement or java.sql.PreparedStatement in hsqldb (HyperSQL DataBase) to process untrusted input may be vulnerable to a remote code execution attack. By default it is allowed to call any static method of any Java class in the classpath resulting in code execution. The issue can be prevented by updating to 2.7.1 or by setting the system property "hsqldb.method_class_names" to classes which are allowed to be called. For example, System.setProperty("hsqldb.method_class_names", "abc") or Java argument -Dhsqldb.method_class_names="abc" can be used. From version 2.7.1 all classes by default are not accessible except those in java.lang.Math and need to be manually enabled.

A sandbox bypass vulnerability in Jenkins Script Security Plugin 1.55 and earlier allows attackers to invoke arbitrary constructors in sandboxed scripts.

A sandbox bypass vulnerability in Jenkins Pipeline: Groovy Plugin 2.64 and earlier allows attackers to invoke arbitrary constructors in sandboxed scripts.

A vulnerability was found in Infinispan such that the invokeAccessibly method from the public class ReflectionUtil allows any application class to invoke private methods in any class with Infinispan's privileges. The attacker can use reflection to introduce new, malicious behavior into the application.

Kylin can receive user input and load any class through Class.forName(...). This issue affects Apache Kylin 2 version 2.6.6 and prior versions; Apache Kylin 3 version 3.1.2 and prior versions; Apache Kylin 4 version 4.0.0 and prior versions.

Emissary is a P2P based data-driven workflow engine. Affected versions of Emissary are vulnerable to post-authentication Remote Code Execution (RCE). The [`CreatePlace`](https://github.com/NationalSecurityAgency/emissary/blob/30c54ef16c6eb6ed09604a929939fb9f66868382/src/main/java/emissary/server/mvc/internal/CreatePlaceAction.java#L36) REST endpoint accepts an `sppClassName` parameter which is used to load an arbitrary class. This class is later instantiated using a constructor with the following signature: `<constructor>(String, String, String)`. An attacker may find a gadget (class) in the application classpath that could be used to achieve Remote Code Execution (RCE) or disrupt the application. Even though the chances to find a gadget (class) that allow arbitrary code execution are low, an attacker can still find gadgets that could potentially crash the application or leak sensitive data. As a work around disable network access to Emissary from untrusted sources.

Sitecore Experience Platform (XP) v9.3 was discovered to contain an authenticated remote code execution (RCE) vulnerability via the component /sitecore/shell/Invoke.aspx.

GLPI is an open-source asset and IT management software package that provides ITIL Service Desk features, licenses tracking and software auditing. In GLPI before version 9.5.4 non-authenticated user can remotely instantiate object of any class existing in the GLPI environment that can be used to carry out malicious attacks, or to start a “POP chain”. As an example of direct impact, this vulnerability affects integrity of the GLPI core platform and third-party plugins runtime misusing classes which implement some sensitive operations in their constructors or destructors. This is fixed in version 9.5.4.

Use of Externally-Controlled Input to Select Classes or Code ('Unsafe Reflection') vulnerability in Mitsubishi Electric Corporation EZSocket versions 3.0 to 5.92, GT Designer3 Version1(GOT1000) versions 1.325P and prior, GT Designer3 Version1(GOT2000) versions 1.320J and prior, GX Works2 versions 1.11M to 1.626C, GX Works3 versions 1.106L and prior, MELSOFT Navigator versions 1.04E to 2.102G, MT Works2 versions 1.190Y and prior, MX Component versions 4.00A to 5.007H and MX OPC Server DA/UA all versions allows a remote unauthenticated attacker to execute a malicious code by RPC with a path to a malicious library while connected to the products.

Avo is an open source ruby on rails admin panel creation framework. The polymorphic field type stores the classes to operate on when updating a record with user input, and does not validate them in the back end. This can lead to unexpected behavior, remote code execution, or application crashes when viewing a manipulated record. This issue has been addressed in commit `ec117882d` which is expected to be included in subsequent releases. Users are advised to limit access to untrusted users until a new release is made.

stimulus_reflex is a system to extend the capabilities of both Rails and Stimulus by intercepting user interactions and passing them to Rails over real-time websockets. In affected versions more methods than expected can be called on reflex instances. Being able to call some of them has security implications. To invoke a reflex a websocket message of the following shape is sent: `\"target\":\"[class_name]#[method_name]\",\"args\":[]`. The server will proceed to instantiate `reflex` using the provided `class_name` as long as it extends `StimulusReflex::Reflex`. It then attempts to call `method_name` on the instance with the provided arguments. This is problematic as `reflex.method method_name` can be more methods that those explicitly specified by the developer in their reflex class. A good example is the instance_variable_set method. This vulnerability has been patched in versions 3.4.2 and 3.5.0.rc4. Users unable to upgrade should: see the backing GHSA advisory for mitigation advice.

Craft CMS is a content management system (CMS). From version 5.6.0 to before version 5.9.13, a Remote Code Execution (RCE) vulnerability exists in Craft CMS, it can be exploited by any authenticated user with control panel access. This is a bypass of a previous fix. The existing patches add cleanseConfig() to assembleLayoutFromPost() and various FieldsController actions to strip Yii2 behavior/event injection keys ("as" and "on" prefixed keys). However, the fieldLayouts parameter in ElementIndexesController::actionFilterHud() is passed directly to FieldLayout::createFromConfig() without any sanitization, enabling the same behavior injection attack chain. This issue has been patched in version 5.9.13.

Spinnaker is an open source, multi-cloud continuous delivery platform. Prior to 2026.1.0, 2026.0.3, 2025.4.4, and 2025.3.3, unsafe YAML processing bypasses safe deserialization when using CloudFormation deployments or CloudFoundry baking. The use of a non-safe constructor allows arbitrary loading of Java classes, leading to remote code execution. This issue is fixed in versions 2026.1.0, 2026.0.3, 2025.4.4, and 2025.3.3.

A vulnerability of XPlatform could allow an unauthenticated attacker to execute arbitrary command. This vulnerability exists due to insufficient validation of improper classes. This issue affects: Tobesoft XPlatform versions prior to 9.2.2.280.

It was found that the fix for CVE-2014-0114 had been reverted in JBoss Operations Network 3 (JON). This flaw allows attackers to manipulate ClassLoader properties on a vulnerable server. Exploits that have been published rely on ClassLoader properties that are exposed such as those in JON 3. Additional information can be found in the Red Hat Knowledgebase article: https://access.redhat.com/site/solutions/869353. Note that while multiple products released patches for the original CVE-2014-0114 flaw, the reversion described by this CVE-2019-3834 flaw only occurred in JON 3.

In pig-mesh Pig versions 3.8.2 and below, when setting up scheduled tasks in the Quartz management function under the system management module, it is possible to execute any Java class with a parameterless constructor and its methods with parameter type String through reflection. At this time, the eval method in Tomcat's built-in class jakarta.el.ELProcessor can be used to execute commands, leading to a remote code execution vulnerability.

Yamcs is a mission control framework. Prior to 5.12.7, the Nashorn ScriptEngine used to evaluate user-supplied JavaScript algorithm text in yamcs-core/src/main/java/org/yamcs/algorithms/ScriptAlgorithmExecutorFactory.java was constructed without a ClassFilter, so a user with the ChangeMissionDatabase privilege could override an algorithm through the MdbOverrideApi.updateAlgorithm endpoint and supply JavaScript that reaches arbitrary Java classes (for example Java.type("java.lang.Runtime").getRuntime().exec(...)) to execute arbitrary OS commands as the Yamcs process; in the default configuration with no security.yaml the built-in guest user has superuser=true, making the issue reachable without authentication. This issue is fixed in versions 5.12.7 and 5.13.0, which disable algorithm editing by default.

Craft is a platform for creating digital experiences. In versions 4.0.0-RC1 through 4.16.17 and 5.0.0-RC1 through 5.8.21, a Remote Code Execution (RCE) vulnerability exists in Craft CMS where the assembleLayoutFromPost() function in src/services/Fields.php fails to sanitize user-supplied configuration data before passing it to Craft::createObject(). This allows authenticated administrators to inject malicious Yii2 behavior configurations that execute arbitrary system commands on the server. This vulnerability represents an unpatched variant of the behavior injection vulnerability addressed in CVE-2025-68455, affecting different endpoints through a separate code path. This vulnerability is fixed in 5.8.22.

codeBeamer before 9.5.0-RC3 does not properly restrict the ability to execute custom Java code and access the Java class loader via computed fields.

In Progress® Telerik® Reporting versions prior to 18.1.24.709, a code execution attack is possible through object injection via an insecure type resolution vulnerability.

Grav versions >= 1.7.0 and before 2.0.9 contain a remote code execution vulnerability. FlexDirectory::dynamicDataField() resolves blueprint data-*@: directives by calling call_user_func_array() on attacker-influenced input, validating only that the target is callable (is_callable()) without restricting dangerous functions such as exec, system, passthru, or shell_exec. Because FlexDirectory registers this handler for every Flex directory, it bypasses the validation added to Blueprint::dynamicData() in 2.0.7 (GHSA-fj2p-qj2f-74v5). Any authenticated user with create or update permission on any Flex-based directory (Flex Users, Flex Pages, Flex Objects, or custom Flex types) can execute arbitrary shell commands on the server.

Craft is a platform for creating digital experiences. Versions 5.0.0-RC1 through 5.8.20 and 4.0.0-RC1 through 4.16.16 are vulnerable to potential authenticated Remote Code Execution via malicious attached Behavior. Note that attackers must have administrator access to the Craft Control Panel for this to work. Users should update to the patched versions (5.8.21 and 4.16.17) to mitigate the issue.

In Progress Telerik Report Server versions prior to 2024 Q3 (10.2.24.924), a remote code execution attack is possible through object injection via an insecure type resolution vulnerability.

A website could have obscured the fullscreen notification by using a URL with a scheme handled by an external program, such as a mailto URL. This could have led to user confusion and possible spoofing attacks. This vulnerability affects Firefox < 115, Firefox ESR < 102.13, and Thunderbird < 102.13.

Barracuda Service Center, as implemented in the RMM solution, in versions prior to 2025.1.1, does not correctly verify the name of an attacker-controlled WSDL service, leading to insecure reflection. This can result in remote code execution through either invocation of arbitrary methods or deserialization of untrusted types.

In Progress® Telerik® UI for AJAX prior to v2026.2.708, forged upload metadata can influence AsyncUploadTypeName processing and trigger unsafe attacker-controlled type resolution, enabling remote code execution in affected deployments.

Arbitrary Class Instantiation via Model Manifest in Apache OpenNLP ExtensionLoader Versions Affected: before 1.9.5, before 2.5.9, before 3.0.0-M3 Description:  The ExtensionLoader.instantiateExtension(Class, String) method loads a class by its fully-qualified name via Class.forName() and invokes its no-arg constructor, with the class name sourced from the manifest.properties entry of a model archive. The existing isAssignableFrom check correctly rejects classes that are not subtypes of the expected extension interface (BaseToolFactory for factory=, ArtifactSerializer for serializer-class-*), but the check runs after Class.forName() has already loaded and initialized the named class. Class.forName() with default initialization semantics executes the target class's static initializer before returning, so an attacker who can supply a crafted model archive can cause the static initializer of any class on the classpath to run during model loading, regardless of whether that class passes the subsequent type check. Exploitation requires a class with attacker-useful side effects in its static initializer (for example, JNDI lookup, outbound network I/O, or filesystem access) to be present on the classpath, so this is not a drop-in remote code execution; however, the attack surface grows as third-party model distribution becomes more common (community model repositories, Hugging Face-style sharing), where users routinely load model files from origins they do not control. A secondary, narrower vector affects deployments that ship legitimate BaseToolFactory or ArtifactSerializer subclasses with side-effecting no-arg constructors: a malicious manifest can name such a class and force its constructor to run during model load. Mitigation:  * 2.x users should upgrade to 2.5.9. * 3.x users should upgrade to 3.0.0-M3. Note: The fix introduces a package-prefix allowlist that is consulted before Class.forName() is invoked, so the static initializer of a disallowed class is never executed. Classes under the opennlp. prefix remain permitted by default. Deployments that load models referencing factories or serializers outside opennlp.* must opt those packages in, either programmatically via ExtensionLoader.registerAllowedPackage(String) before the first model load, or by setting the OPENNLP_EXT_ALLOWED_PACKAGES system property to a comma-separated list of allowed package prefixes. Users who cannot upgrade immediately should ensure that all model files are sourced from trusted origins and should audit their classpath for classes with side-effecting static initializers or constructors, particularly any that perform JNDI lookups, network requests, or filesystem operations during class initialization.

ColdFusion MX 6.1 and 6.1 J2EE allows local users to bypass sandbox security restrictions and obtain sensitive information by using Java reflection methods to access trusted Java objects without using the CreateObject function or cfobject tag.

Remote Code Execution via Arbitrary Class Instantiation in plugin-schema-registry component in Apache Ranger <= 2.8.0. Users are recommended to upgrade to version 2.9.0, which fixes this issue.

IdentityIQ 8.3 and all 8.3 patch levels prior to 8.3p3, IdentityIQ 8.2 and all 8.2 patch levels prior to 8.2p6, IdentityIQ 8.1 and all 8.1 patch levels prior to 8.1p7, IdentityIQ 8.0 and all 8.0 patch levels prior to 8.0p6 allow an authenticated user to invoke a Java constructor with no arguments or a Java constructor with a single Map argument in any Java class available in the IdentityIQ application classpath.

In Progress Telerik Reporting versions prior to 2024 Q3 (18.2.24.924), a code execution attack is possible through object injection via an insecure type resolution vulnerability.

Use of Externally-Controlled Input to Select Classes or Code ('Unsafe Reflection') vulnerability in Apache IoTDB. The pipe processor reads a fully qualified Java class name and instantiates it using Class.forName().newInstance() without any validation or allowlisting. This issue affects Apache IoTDB: from 1.0.0 before 2.0.10. Users are recommended to upgrade to version 2.0.10, which fixes the issue.

An issue exists in Amazon Redshift JDBC Driver versions prior to 2.2.2. Under certain conditions, the driver could load and execute arbitrary classes when processing JDBC connection URL parameters. An actor who can influence the connection URL could potentially execute code in the application context, provided a suitable class is available on the application's classpath. To mitigate this issue, users should upgrade to version 2.2.2 or later.

Arbitrary Class Instantiation via XML Feature Generator Descriptor and Format Name in Apache OpenNLP Versions Affected: - before 2.5.10 - before 3.0.0-M5 Description: Three code paths in Apache OpenNLP load a class by its fully-qualified name via Class.forName() and invoke its no-arg constructor without any prior validation of the class name or its type.  The affected paths are: (1) GeneratorFactory, which reads the class attribute of generator elements in an XML feature generator descriptor; such descriptors are embedded as artifacts in model archives (e.g. TokenNameFinder and POSTagger models) and are parsed during model loading, so an attacker who can supply a crafted model archive controls the class name directly. (2) StreamFactoryRegistry.getFactory(Class, String), which falls back to interpreting an unregistered format name as the fully-qualified class name of an ObjectStreamFactory; this is exploitable in applications that pass untrusted format names (e.g. exposing the -format parameter of the command-line tooling to external input). (3) StringInterners, which instantiates the interner implementation named by the opennlp.interner.class system property; this value is normally deployer-controlled, so it is hardened as defense in depth rather than being independently attacker-reachable. Exploitation requires a class with attacker-useful side effects in its static initializer or no-arg constructor (JNDI lookup, outbound network I/O, filesystem access) to be present on the classpath, so this is not drop-in remote code execution. T Mitigation: Upgrade to a fixed release. The fix routes all three paths through ExtensionLoader.instantiateExtension(...), which consults a package-prefix allowlist before Class.forName() is invoked, so a disallowed class is never loaded, initialized, or constructed. Classes under the opennlp. prefix remain permitted by default. Deployments that load models referencing feature generator factories, object stream factories, or string interners outside opennlp.* must opt those packages in, either programmatically via ExtensionLoader.registerAllowedPackage(String) before the first model load, or by setting the OPENNLP_EXT_ALLOWED_PACKAGES system property to a comma-separated list of allowed package prefixes. Users who cannot upgrade immediately should ensure all model files and format names are sourced from trusted origins and should audit their classpath for classes with side-effecting static initializers or constructors.

Yamcs is a mission control framework. Prior to 5.12.8 and 5.13.2, Yamcs inserts templateArgs from POST /api/instances and PATCH /api/instances/{instance} into YAML through VarStatement.append in yamcs-core/src/main/java/org/yamcs/templating/VarStatement.java without YAML-context escaping. The rendered configuration is parsed by YamcsServer.createInstance and loaded by YamcsServerInstance, allowing an attacker to inject a services entry for org.yamcs.ProcessRunner. Deployments without security.yaml expose the operation through the guest superuser, while secured deployments require SystemPrivilege.CreateInstances. Successful exploitation executes commands as the Yamcs service account. This issue is fixed in versions 5.12.8 and 5.13.2.

A high-severity vulnerability that can lead to arbitrary code execution on the system hosting the Web SDK role was found in the Genetec Security Center product line.

CtrlPanel is open-source billing software for hosting providers. In versions 1.1.1 and prior, the admin settings update endpoint accepted a fully qualified class name directly from user-supplied request input and used it for dynamic static method calls and object instantiation without any allowlist validation, allowing for authenticated Remote Code Execution. An authenticated admin-level user could supply an arbitrary class name available in the Composer autoloader, potentially triggering unintended constructor or magic method execution. The update() method reads settings_class directly from the HTTP request and passed it to new $settings_class() and $settings_class::getValidations() without verifying that the provided value corresponds to a legitimate settings class: Because PHP resolves class names against the Composer autoloader at runtime, any autoloadable class in the application or its dependencies could be instantiated. Depending on the classes available in the dependency tree, this can trigger unintended side effects through constructors or magic methods (__construct, __toString, __wakeup), following a PHP object injection / gadget chain pattern. This issue has been fixed in version 1.2.0.

generator-jhipster-entity-audit is a JHipster module to enable entity audit and audit log page. Prior to 5.9.1, generator-jhipster-entity-audit allows unsafe reflection when having Javers selected as Entity Audit Framework. If an attacker manages to place some malicious classes into the classpath and also has access to these REST interface for calling the mentioned REST endpoints, using these lines of code can lead to unintended remote code execution. This vulnerability is fixed in 5.9.1.

Spring Authorization Server versions 1.0.0 - 1.0.5, 1.1.0 - 1.1.5, 1.2.0 - 1.2.2 and older unsupported versions are susceptible to a PKCE Downgrade Attack for Confidential Clients. Specifically, an application is vulnerable when a Confidential Client uses PKCE for the Authorization Code Grant. An application is not vulnerable when a Public Client uses PKCE for the Authorization Code Grant.

The ShopLentor plugin for WordPress is vulnerable to arbitrary function execution via the woolentoropt/v1/custom-action REST API endpoint in all versions up to, and including, 3.3.7. This is due to the handle_action() method passing user-supplied input directly to call_user_func() without an allowlist of permitted callbacks. This makes it possible for authenticated attackers, with Administrator-level access and above, to execute arbitrary PHP callable functions via the 'callback' parameter.

DBI versions before 1.650 for Perl are vulnerable to code injection via caller-influenced Profile. When a string is assigned to a DBI handle's Profile attribute, DBI splits it into path, package and arguments, and interpolates the package part in a string eval with no validation of the package name. Any caller-influenced value that reaches the Profile attribute is therefore arbitrary Perl code execution, including calls to run system commands. The Profile attribute can be set from three different sources that can carry untrusted data: the DBI_PROFILE environment variable, a direct attribute assignment, and a DSN driver-attribute clause dbi:Driver(Profile=>SPEC):db. An attacker controlling any of those inputs runs arbitrary Perl in the host process. The strongest remote position is a network-exposed DBI::Gofer / DBI::ProxyServer whose per-request DSN reaches the Profile attribute, letting a client execute code on the broker host.

Craft CMS is a content management system (CMS). From version 4.0.0-RC1 to before version 4.17.5 and from version 5.0.0-RC1 to before version 5.9.11, there is a Behavior injection RCE vulnerability in ElementIndexesController and FieldsController. Craft control panel administrator permissions and allowAdminChanges must be enabled for this to work. This issue has been patched in versions 4.17.5 and 5.9.11.

IBM WebSphere Extreme Scale 8.6.1.0 through 8.6.1.6 's Object Query Language engine resolves attacker-supplied class names via Class.forName() and invokes their constructors with no allow-list at three distinct sinks (SELECT NEW, enum literals, and reflection-based comparators); an authenticated remote attacker who can influence an application-built OQL query string can execute arbitrary constructors on the WAS JVM, and a SELECT DISTINCT variant using planted grid values fires the same gadget post-readObject in a manner that survives JEP-290 serialization filters across grid node boundaries

The Addressing GLPI plugin enables you to create IP reports for visualize IP addresses used and free on a given network.. Starting with 3.0.0 and before 3.0.3, a poor security check allows an unauthenticated attacker to determine whether data exists (by name) in GLPI.

Craft CMS is a content management system (CMS). From version 5.6.0 to before version 5.9.11, in src/controllers/EntryTypesController.php, the $settings array from parse_str is passed directly to Craft::configure() without Component::cleanseConfig(). This allows injecting Yii2 behavior/event handlers via "as" or "on" prefixed keys, the same attack vector as the original advisory. Craft control panel administrator permissions and allowAdminChanges must be enabled for this to work. This issue has been patched in version 5.9.11.

Kirby is an open-source content management system. Prior to 4.9.1 and 5.4.1, Kirby did not validate the model attributes that were used in its collection queries, allowing attackers to include arbitrary model methods in their queries. This includes methods with sensitive data such as password() (disclosing the password hash) or root() (disclosing the absolute filesystem path on the server) as well as methods that perform impactful actions such as loginPasswordless() (causing a privilege escalation to another user) or delete() (deleting all queried models in one go if the authenticated user has appropriate permissions). This issue has been fixed in versions 4.9.1 and 5.4.1.

HTML::FormHandler versions through 0.40068 for Perl allow attacker selected method dispatch and resource exhaustion because _apply_actions and add_error use error message text built from request data as a Locale::Maketext bracket notation template. add_error hands its first argument to the language handle as the Locale::Maketext message key, and the default handle's lexicon sets `_AUTO`, so a string that is not a lexicon entry is compiled as a bracket notation template instead of being looked up. In a bracket group the first token names a method called on the language handle and the remaining tokens are its arguments. Three kinds of text the library did not author reach that position. _apply_actions installs a `$SIG{__WARN__}` handler that stores the warning text in `$error_message`, and a captured warning survives a successful action, so a field carrying a numeric transform turns `Argument "[sprintf,%50000000d,0]" isn't numeric` into the template; a warning quotes the submitted value verbatim, so the group is well formed and dispatches. `$error_message ||= $tobj->validate($new_value)` takes a type constraint's own failure message, which renders the rejected value through a partial dumper in bracket and comma form (Devel::PartialDump when Moose can load it, Type::Tiny's own dumper always), so a field with `apply => [ Str ]` given a parameter sent more than once, which arrives as an array, gets `Reference ["a","b"] did not pass type constraint "Str"` as its template, from a request that carries no bracket character of its own. A coercion or transform exception reaches it the same way. Beyond those, a validator whose message contains the field value puts that value in the template directly, and add_error replaces the message list with the contents of an arrayref first argument (`@message = @{$message[0]} if ref $message[0] eq 'ARRAY'`), so a value arriving as an array fills the argument slots from the same request as well. A malformed group such as `[0]` makes the compile croak, and HTML::FormHandler::I18N::maketext and add_error each re-raise that as a die, so process() throws. A well formed group naming sprintf reaches CORE::sprintf with an attacker chosen field width. Any caller that applies a type constraint or a transform to an untrusted field, or whose validator passes an untrusted field value to add_error, can be made to throw an unhandled exception out of process(), or to allocate an arbitrary amount of memory in one request, and an application whose language handle subclass defines side effecting public methods makes those callable with attacker chosen arguments. The dumped type constraint message is bounded to the exception, because both dumpers quote non-numeric elements so the method slot is never an attacker chosen name. The built-in messages pass fixed templates with the value in an argument slot, where it stays inert, and the built-in field types attach explicit message callbacks, so neither is affected.

In Hibernate Validator 5.2.x before 5.2.5 final, 5.3.x, and 5.4.x, it was found that when the security manager's reflective permissions, which allows it to access the private members of the class, are granted to Hibernate Validator, a potential privilege escalation can occur. By allowing the calling code to access those private members without the permission an attacker may be able to validate an invalid instance and access the private member value via ConstraintViolation#getInvalidValue().

In Progress® Telerik® UI for AJAX prior to v2026.2.708, DialogHandler provider type input may be tampered with, potentially altering dialog processing and enabling chained exploitation.

An unsafe dynamic class loading vulnerability exists in the database connection utilities of PaperCut MF and PaperCut NG. The application instantiates database driver classes based on configurable driver names without validating against an allowlist of approved drivers. If an attacker can manipulate system configuration parameters, this enables the execution of arbitrary Java bytecode residing on the application classpath under the security context of the PaperCut server process.

Statamic is a Laravel and Git powered content management system (CMS). Prior to 5.73.23 and 6.20.0, the fix for CVE-2026-41175 was incomplete. It addressed the issue in the query builder, but the same protection was not applied to in-memory collection sorting. Manipulating sort parameters could result in the loss of content and assets. This requires a front-end template that passes request input into a tag's sort parameter. It is not exploitable by default — a template would need to be explicitly set up to sort by a visitor-controlled value. This has been fixed in 5.73.23 and 6.20.0.

An unsafe reflection vulnerability in Kentico Xperience allows an unauthenticated attacker to kill the current process, leading to a Denial-of-Service condition. This issue affects Xperience: through 13.0.180.

Form::Processor::Field::HtmlArea versions from 0.06 through 1.162360 for Perl allow attacker selected method dispatch and resource exhaustion via an HTML::Tidy diagnostic that validate passes to add_error as a Locale::Maketext template. validate runs HTML::Tidy over the submitted markup and passes each resulting message to add_error as its first argument, which add_error hands to the language handle as the Locale::Maketext message key. The default handle's lexicon sets `_AUTO`, so a message that is not a lexicon entry is compiled as a bracket notation template instead of being looked up. Tidy diagnostics quote the offending attribute name or value, so a bracket group in the submitted markup reaches the template position, where the first token of the group names a method called on the language handle and the remaining tokens are its arguments. A group such as `[0]` makes the compile croak, and neither the field nor the handle catches it, so the exception leaves validate. `[sprintf,%2000000000d,7]` reaches CORE::sprintf with an attacker chosen field width. One submission of crafted markup to an HtmlArea field throws an unhandled exception out of form validation or allocates an arbitrary amount of memory, and an application whose language handle subclass defines side effecting public methods makes those callable with attacker chosen arguments. The other field types pass fixed templates with the submitted value in an argument slot, where it stays inert, and are unaffected.

An issue in AWS Wrappers for Amazon Aurora PostgreSQL may allow for privilege escalation to rds_superuser role. A low privilege authenticated user can create a crafted function that could be executed with permissions of other Amazon Relational Database Service (RDS) users. We recommend customers upgrade to the following versions: AWS JDBC Wrapper to v2.6.5, AWS Go Wrapper to 2025-10-17, AWS NodeJS Wrapper to v2.0.1, AWS Python Wrapper to v1.4.0 and AWS PGSQL ODBC driver to v1.0.1

MessagePack for C# is a MessagePack serializer for C#. Prior to 2.5.301 and 3.1.7, MessagePackReader.ReadDateTime() can allocate stack memory based on an attacker-controlled MessagePack extension length. In the slow path for timestamp extension parsing, the computed tokenSize includes the extension body length from the wire and is used in a stackalloc operation before the extension length is validated as one of the valid timestamp sizes. A very small payload can claim a large timestamp extension body and cause a stack allocation large enough to trigger an uncatchable StackOverflowException, terminating the host process. This vulnerability is fixed in 2.5.301 and 3.1.7.

Use of Externally-Controlled Input to Select Classes or Code ('Unsafe Reflection') vulnerability in Apache Calcite. This issue affects Apache Calcite: from 1.5.0 before 1.42. Users are recommended to upgrade to version 1.42, which fixes the issue.

NVIDIA Megatron Bridge for Linux contains a vulnerability where an attacker could cause improper control of dynamically managed code resources. A successful exploit of this vulnerability might lead to code execution, escalation of privileges, data tampering, and information disclosure.

Astro is a web framework. Prior to version 5.14.2, Astro reflects the value in `X-Forwarded-Host` in output when using `Astro.url` without any validation. It is common for web servers such as nginx to route requests via the `Host` header, and forward on other request headers. As such as malicious request can be sent with both a `Host` header and an `X-Forwarded-Host` header where the values do not match and the `X-Forwarded-Host` header is malicious. Astro will then return the malicious value. This could result in any usages of the `Astro.url` value in code being manipulated by a request. For example if a user follows guidance and uses `Astro.url` for a canonical link the canonical link can be manipulated to another site. It is theoretically possible that the value could also be used as a login/registration or other form URL as well, resulting in potential redirecting of login credentials to a malicious party. As this is a per-request attack vector the surface area would only be to the malicious user until one considers that having a caching proxy is a common setup, in which case any page which is cached could persist the malicious value for subsequent users. Many other frameworks have an allowlist of domains to validate against, or do not have a case where the headers are reflected to avoid such issues. This could affect anyone using Astro in an on-demand/dynamic rendering mode behind a caching proxy. Version 5.14.2 contains a fix for the issue.

PraisonAI is a multi-agent teams system. Prior to praisonai version 4.6.37 and praisonaiagents version 1.6.37, praisonaiagents resolves unresolved tool names against module globals and __main__ after it fails to match the declared tool list and the registry. With the default agent configuration, _perm_allow is None, so undeclared non-dangerous tool names are not rejected by the permission gate. An attacker who can influence tool-call names can therefore invoke unintended application callables that were never declared as tools. This issue has been patched in praisonai version 4.6.37 and praisonaiagents version 1.6.37.

Craft CMS is a content management system (CMS). From 4.0.0 to before 4.17.12 and 5.9.18, Craft CMS which contains an input-handling flaw in a Yii object creation path that let any authenticated user inject malicious configuration and execute arbitrary commands on the server. The request-controlled condition field layouts data is converted into a live FieldLayout object without a Component::cleanseConfig() boundary. Because Craft configures models before parent::__construct(), attacker-controlled special config keys can take effect during object creation, and FieldLayout initialization then triggers a same-request event. This vulnerability is fixed in 4.17.12 and 5.9.18.

MessagePack for C# is a MessagePack serializer for C#. Prior to 2.5.301 and 3.1.7, MessagePack-CSharp's typeless deserialization includes MessagePackSerializerOptions.ThrowIfDeserializingTypeIsDisallowed(Type) as a safety check for dangerous types. The default implementation checks the outer type name, but it does not recursively inspect array element types or generic type arguments. As a result, a type that would be blocked directly can be wrapped inside an array or constructed generic type and pass the outer type check. The formatter machinery can then materialize formatters for the inner blocked type. This vulnerability is fixed in 2.5.301 and 3.1.7.

Jenkins Pipeline: Groovy Plugin 4331.v9d06ed4658ff and earlier does not restrict the types that can be instantiated through the Pipeline Snippet Generator, allowing attackers to instantiate types related to job or system configuration other than Pipeline steps.

PyTorch Lightning through 2.6.5, fixed in commit d710d68, contains a remote code execution vulnerability in the _load_state function that imports and executes attacker-controlled module names from checkpoint _instantiator hyperparameters. Attackers can craft malicious checkpoint files that bypass weights_only=True protections to execute arbitrary code when LightningModule.load_from_checkpoint is called.

Vocos instantiates a class named by a configuration file without restricting which class may be named. instantiate_class in vocos/pretrained.py takes the class_path value from the configuration, splits it into a module and an attribute, imports the module with __import__, resolves the attribute with getattr, and calls the result as args_class(*args, **kwargs) where kwargs is the config's own init_args mapping. No allowlist constrains the dotted path, so a configuration may name any importable callable and supply the arguments it is called with. Vocos.from_hparams reaches this for each of the feature_extractor, backbone and head entries, and Vocos.from_pretrained reaches it with a remote file: it downloads config.yaml from a caller-named Hugging Face repository and passes it straight to from_hparams. Loading a model from a repository the user does not control therefore executes code of the repository owner's choosing in the loading process. The neighbouring torch.load of the downloaded weights is a separate matter and is constrained on PyTorch releases that default weights_only to true, which leaves this path as the reachable one.

The RabbitMQ Java client library allows Java and JVM-based applications to connect to and interact with RabbitMQ nodes. Prior to 5.33.0, com.rabbitmq.tools.jsonrpc.ProcedureDescription receives a javaReturnType value in an untrusted system.describe response and passes it through JSONUtil.tryFill, setJavaReturnType, and computeReturnTypeAsJavaClass to Class.forName(javaReturnType) with initialization enabled. An attacker able to answer the JsonRpcClient request through a shared broker or network interception can select a class already present in the victim JVM and trigger its static initializer, while JsonRpcClient.java later passes getReturnType output to mapper.parse and may also create type confusion. Successful exploitation can affect confidentiality, integrity, and availability in the client process. This issue is fixed in version 5.33.0.

DebugKit provides a debugging toolbar for CakePHP applications. Prior to 4.10.3 and 5.2.4, the DebugKit MailPreview feature in src/Controller/MailPreviewController.php accepts a route-controlled previewName value in findPreview and passes the resolved class from App::className() to constructor execution without rejecting namespace separators or verifying that the class extends DebugKit\Mailer\MailPreview. An attacker able to access DebugKit while debug mode is enabled and the request hostname is local or allowlisted can select an unintended application class through the mail-preview preview route, resulting in arbitrary constructor execution and limited disclosure of application information. This issue is fixed in versions 4.10.3 and 5.2.4.