Deserialization of Untrusted Data in sagemaker-python-sdk

sagemaker-python-sdk is a library for training and deploying machine learning models on Amazon SageMaker. The sagemaker.base_deserializers.NumpyDeserializer module before v2.218.0 allows potentially unsafe deserialization when untrusted data is passed as pickled object arrays. This consequently may allow an unprivileged third party to cause remote code execution, denial of service, affecting both confidentiality and integrity. Users are advised to upgrade to version 2.218.0. Users unable to upgrade should not pass pickled numpy object arrays which originated from an untrusted source, or that could have been tampered with. Only pass pickled numpy object arrays from trusted sources.

Command Injection in sagemaker-python-sdk

sagemaker-python-sdk is a library for training and deploying machine learning models on Amazon SageMaker. In affected versions the capture_dependencies function in `sagemaker.serve.save_retrive.version_1_0_0.save.utils` module allows for potentially unsafe Operating System (OS) Command Injection if inappropriate command is passed as the “requirements_path” parameter. This consequently may allow an unprivileged third party to cause remote code execution, denial of service, affecting both confidentiality and integrity. This issue has been addressed in version 2.214.3. Users are advised to upgrade. Users unable to upgrade should not override the “requirements_path” parameter of capture_dependencies function in `sagemaker.serve.save_retrive.version_1_0_0.save.utils`, and instead use the default value.

D-Tale allows Remote Code Execution through the Custom Filter Input

D-Tale is a visualizer for pandas data structures. Prior to version 3.16.1, users hosting D-Tale publicly can be vulnerable to remote code execution allowing attackers to run malicious code on the server. Users should upgrade to version 3.16.1 where the `update-settings` endpoint blocks the ability for users to update the `enable_custom_filters` flag. The only workaround for versions earlier than 3.16.1 is to only host D-Tale to trusted users.

An SEH-based buffer overflow in the BPQ32 HTTP Server in BPQ32 6.0.24.1 allows remote attackers with access to the Web Terminal to achieve remote code execution via an HTTP POST /TermInput request.

WP STAGING WordPress Backup Plugin – Migration Backup Restore <= 3.4.3 - Authenticated (Admin+) Arbitrary File Upload

The WP STAGING WordPress Backup Plugin – Migration Backup Restore plugin for WordPress is vulnerable to arbitrary file uploads due to missing file type validation in the wpstg_processing AJAX action in all versions up to, and including, 3.4.3. This makes it possible for authenticated attackers, with administrator-level access and above, to upload arbitrary files on the affected site's server which may make remote code execution possible.

Prompt Injection in "ask" API with visualization leads to RCE

The Vanna library uses a prompt function to present the user with visualized results, it is possible to alter the prompt using prompt injection and run arbitrary Python code instead of the intended visualization code. Specifically - allowing external input to the library’s “ask” method with "visualize" set to True (default behavior) leads to remote code execution.

Invoice Ninja before 5.10.43 allows remote code execution from a pre-authenticated route when an attacker knows the APP_KEY. This is exacerbated by .env files, available from the product's repository, that have default APP_KEY values. The route/{hash} route defined in the invoiceninja/routes/client.php file can be accessed without authentication. The parameter {hash} is passed to the function decrypt that expects a Laravel ciphered value containing a serialized object. (Furthermore, Laravel contains several gadget chains usable to trigger remote command execution from arbitrary deserialization.) Therefore, an attacker in possession of the APP_KEY is able to fully control a string passed to an unserialize function.

llama-cpp-python vulnerable to Remote Code Execution by Server-Side Template Injection in Model Metadata

llama-cpp-python is the Python bindings for llama.cpp. `llama-cpp-python` depends on class `Llama` in `llama.py` to load `.gguf` llama.cpp or Latency Machine Learning Models. The `__init__` constructor built in the `Llama` takes several parameters to configure the loading and running of the model. Other than `NUMA, LoRa settings`, `loading tokenizers,` and `hardware settings`, `__init__` also loads the `chat template` from targeted `.gguf` 's Metadata and furtherly parses it to `llama_chat_format.Jinja2ChatFormatter.to_chat_handler()` to construct the `self.chat_handler` for this model. Nevertheless, `Jinja2ChatFormatter` parse the `chat template` within the Metadate with sandbox-less `jinja2.Environment`, which is furthermore rendered in `__call__` to construct the `prompt` of interaction. This allows `jinja2` Server Side Template Injection which leads to remote code execution by a carefully constructed payload.

A PHP Code Injection vulnerability that can lead to Remote Code Execution (RCE) and XSS in Opencode Mobile Collect Call v5.4.7 allows attackers to execute arbitrary web scripts or HTML via a crafted payload injected into the op_func parameter at /occontrolpanel/index.php.

A remote code execution (RCE) vulnerability in the ZScript function of ZDoom Team GZDoom v4.13.1 allows attackers to execute arbitrary code via supplying a crafted PK3 file containing a malicious ZScript source file.

libxmljs2 attrs type confusion RCE

libxmljs2 is vulnerable to a type confusion vulnerability when parsing a specially crafted XML while invoking a function on the result of attrs() that was called on a parsed node. This vulnerability might lead to denial of service (on both 32-bit systems and 64-bit systems), data leak, infinite loop and remote code execution (on 32-bit systems with the XML_PARSE_HUGE flag enabled).

libxmljs2 namespaces type confusion RCE

libxmljs2 is vulnerable to a type confusion vulnerability when parsing a specially crafted XML while invoking the namespaces() function (which invokes XmlNode::get_local_namespaces()) on a grand-child of a node that refers to an entity. This vulnerability can lead to denial of service and remote code execution.

A JNDI injection issue was discovered in Cloudera JDBC Connector for Hive before 2.6.26 and JDBC Connector for Impala before 2.6.35. Attackers can inject malicious parameters into the JDBC URL, triggering JNDI injection during the process when the JDBC Driver uses this URL to connect to the database. This could lead to remote code execution. JNDI injection is possible via the JDBC connection property krbJAASFile for the Java Authentication and Authorization Service (JAAS). Using untrusted parameters in the krbJAASFile and/or remote host can trigger JNDI injection in the JDBC URL through the krbJAASFile.

Remote Code Execution via Path Traversal in parisneo/lollms

CVE-2024-4320 describes a vulnerability in the parisneo/lollms software, specifically within the `ExtensionBuilder().build_extension()` function. The vulnerability arises from the `/mount_extension` endpoint, where a path traversal issue allows attackers to navigate beyond the intended directory structure. This is facilitated by the `data.category` and `data.folder` parameters accepting empty strings (`""`), which, due to inadequate input sanitization, can lead to the construction of a `package_path` that points to the root directory. Consequently, if an attacker can create a `config.yaml` file in a controllable path, this path can be appended to the `extensions` list and trigger the execution of `__init__.py` in the current directory, leading to remote code execution. The vulnerability affects versions up to 5.9.0, and has been addressed in version 9.8.

Authenticated Remote Code Execution (RCE) in HPE Aruba Networking AirWave Management Platform

An authenticated Remote Code Execution (RCE) vulnerability exists in the AirWave CLI. Successful exploitation of this vulnerability could allow a remote authenticated threat actor to run arbitrary commands as a privileged user on the underlying host.

A remote code execution (RCE) vulnerability in Arcadyan Meteor 2 CPE FG360 Firmware ETV2.10 allows attackers to execute arbitrary code via a crafted request.

An Authenticated Remote Code Execution (RCE) vulnerability affects the TP-Link Archer router series. A vulnerability exists in the "tmp_get_sites" function of the HomeShield functionality provided by TP-Link. This vulnerability is still exploitable without the activation of the HomeShield functionality.

Remote Code Execution through Deserialization of Untrusted Data in Splunk Secure Gateway app

In Splunk Enterprise versions below 9.3.2, 9.2.4, and 9.1.7, and versions below 3.4.261 and 3.7.13 of the Splunk Secure Gateway app on Splunk Cloud Platform, a low-privileged user that does not hold the “admin“ or “power“ Splunk roles could perform a Remote Code Execution (RCE).

Potential XXE (XML External Entity Injection) vulnerability in veraPDF CLI

veraPDF is an open source PDF/A validation library. Executing policy checks using custom schematron files via the CLI invokes an XSL transformation that may theoretically lead to a remote code execution (RCE) vulnerability. This doesn't affect the standard validation and policy checks functionality, veraPDF's common use cases. Most veraPDF users don't insert any custom XSLT code into policy profiles, which are based on Schematron syntax rather than direct XSL transforms. For users who do, only load custom policy files from sources you trust. This issue has not yet been patched. Users are advised to be cautious of XSLT code until a patch is available.

ReDoS in Giskard Scan text perturbation

Giskard is an evaluation and testing framework for AI systems. A Remote Code Execution (ReDoS) vulnerability was discovered in Giskard component by the GitHub Security Lab team. When processing datasets with specific text patterns with Giskard detectors, this vulnerability could trigger exponential regex evaluation times, potentially leading to denial of service. Giskard versions prior to 2.15.5 are affected.