Kasimir Schulz

Name of the Vulnerable Software and Affected Versions: MindsDB (affected versions not specified) Description: A cross-site scripting (XSS) issue exists, enabling the execution of a JavaScript payload whenever a user enumerates an ML Engine, database, project, or dataset containing arbitrary JavaScript code within the web UI. This allows a JavaScript payload to run when a user views certain elements in the web UI. Recommendations: At the moment, there is no information about a newer version that contains a fix for this vulnerability.

An arbitrary code execution issue exists due to the way the Refuel Autolabel library handles provided CSV files in its classification tasks. If a maliciously crafted CSV file containing Python code is used to create a classification task, the code will be executed by an `eval` function.

An arbitrary code execution issue exists due to the way multilabel classification tasks handle provided CSV files. If a user creates a multilabel classification task using a maliciously crafted CSV file containing Python code, the code will be passed to an `eval` function which executes it.

Name of the Vulnerable Software and Affected Versions: Cleanlab versions 2.4.0 or newer Description: The issue is related to the deserialization of untrusted data, which can occur in the Cleanlab project. This allows a maliciously crafted `datalab.pkl` file to run arbitrary code on an end user's system when the data directory is loaded. Recommendations: For versions 2.4.0 or newer, consider avoiding the use of untrusted data in the deserialization process until a patch is available. As a temporary workaround, restrict access to the data directory to minimize the risk of exploitation. At the moment, there is no information about a newer version that contains a fix for this issue.

Name of the Vulnerable Software and Affected Versions: MindsDB versions 23.10.2.0 and newer Description: The issue concerns the deserialization of untrusted data in the MindsDB platform. This allows a maliciously uploaded 'inhouse' model to run arbitrary code on the server when used for a prediction. Recommendations: For versions 23.10.2.0 and newer, consider disabling the deserialization of untrusted data until a patch is available. Restrict access to the 'inhouse' model upload feature to minimize the risk of exploitation. Avoid using the affected versions for prediction tasks that involve untrusted models. At the moment, there is no information about a newer version that contains a fix for this vulnerability.

Name of the Vulnerable Software and Affected Versions: MindsDB versions 23.10.3.0 and newer Description: The issue is related to the deserialization of untrusted data in the MindsDB platform. This allows a maliciously uploaded 'inhouse' model to run arbitrary code on the server when a 'describe' query is run on it. Recommendations: For MindsDB versions 23.10.3.0 and newer, consider disabling the deserialization of untrusted data as a temporary workaround until a patch is available. Restrict access to the 'describe' query functionality to minimize the risk of exploitation. At the moment, there is no information about a newer version that contains a fix for this vulnerability.

Name of the Vulnerable Software and Affected Versions: MindsDB versions 23.10.2.0 and newer Description: The issue concerns the deserialization of untrusted data in the MindsDB platform. This allows a maliciously uploaded 'inhouse' model to run arbitrary code on the server when using 'finetune' on it. The vulnerability is remotely exploitable. Recommendations: For MindsDB versions 23.10.2.0 and newer, consider disabling the 'finetune' functionality for 'inhouse' models until a patch is available to prevent arbitrary code execution. Restrict the upload of 'inhouse' models to trusted sources to minimize the risk of exploitation. As a temporary workaround, review and enhance the security posture of the MindsDB platform to prevent deserialization of untrusted data.

An arbitrary code execution issue exists when the Microsoft SharePoint integration is installed on the server. For databases created with the SharePoint engine, a specially crafted `INSERT` query containing Python code can be used to execute code on the server. This occurs because the code is passed to an `eval` function. The vulnerability is related to incorrect code generation management in the `eval` function of the MindsDB platform, allowing a remote attacker to execute arbitrary code by injecting a specially crafted `INSERT` query.

The issue allows deserialization of untrusted data, enabling a maliciously crafted model to run arbitrary code on an end user's system when loaded.

The R statistical programming language has a deserialization vulnerability that can occur when untrusted data is deserialized, allowing a maliciously crafted RDS (R Data Serialization) formatted file or R package to run arbitrary code on an end user's system when interacted with. This vulnerability can be exploited through the use of specially crafted RDS files or R packages, potentially exposing projects to supply chain attacks. The issue is related to the use of promise objects and lazy evaluation in R.

A vulnerability in the Cisco IOx application hosting environment could allow an authenticated, remote attacker to execute arbitrary commands as root on the underlying host operating system. This issue is due to incomplete sanitization of parameters passed for application activation. An attacker could exploit this by deploying and activating an application with a crafted activation payload file, potentially allowing the execution of arbitrary commands as root.