Redyank

**Name of the Vulnerable Software and Affected Versions** gdown versions prior to 5.2.2 **Description** A Path Traversal issue exists within the `extractall()` function in the `gdown/extractall.py` file. The software fails to sanitize or validate the filenames of members within ZIP or TAR archives during extraction. This allows files to be written outside the intended destination directory, which can lead to arbitrary file overwrite and Remote Code Execution (RCE). **Recommendations** Update to version 5.2.2. As a temporary workaround, restrict the use of the `extractall()` function when processing archives from untrusted sources.

Name of the Vulnerable Software and Affected Versions FastFeedParser versions prior to 0.5.10 Description FastFeedParser, a high-performance RSS, Atom, and RDF parser, is susceptible to a denial-of-service issue. When the `parse()` function fetches a URL that redirects via an HTML `<meta http-equiv="refresh">` tag, it can enter an infinite recursion loop if the redirect chain is unbounded. This occurs due to a lack of depth limiting, visited URL deduplication, and redirect count capping. An attacker controlling the server can exhaust the Python call stack, leading to a process crash. This issue can also be combined with a Server-Side Request Forgery (SSRF) issue to access internal network targets. Recommendations Update to FastFeedParser version 0.5.10 or later.

Name of the Vulnerable Software and Affected Versions Lupa versions 2.6 and earlier Description Lupa integrates the runtimes of Lua or LuaJIT2 into CPython. In versions 2.6 and earlier, the `attribute filter` is not consistently applied when attributes are accessed through built-in functions like `getattr` and `setattr`. This allows an attacker to bypass intended restrictions and potentially achieve arbitrary code execution. Recommendations Update to a version of Lupa later than 2.6.

Name of the Vulnerable Software and Affected Versions Kedro-Datasets versions prior to 9.3.0 Description The Kedro-Datasets plugin is susceptible to a path traversal issue in the PartitionedDataset component. Partition IDs were directly concatenated with the dataset base path without proper validation. Malicious input containing '..' components within a partition ID could allow an attacker to write files outside the intended dataset directory, potentially overwriting arbitrary files on the filesystem. This affects users of PartitionedDataset with any storage backend (local filesystem, S3, GCS, etc.). Recommendations Upgrade to kedro-datasets version 9.3.0 or later. As a temporary workaround, validate partition IDs before passing them to PartitionedDataset, ensuring they do not contain '..' path components.

Name of the Vulnerable Software and Affected Versions curl cffi (affected versions not specified) Description curl cffi does not restrict requests to internal IP ranges and automatically follows redirects via libcurl. This allows an attacker-controlled URL to redirect requests to internal services, such as cloud metadata endpoints. Additionally, curl cffi’s TLS impersonation feature can make these requests appear as legitimate browser traffic, potentially bypassing network controls. The issue arises because user-supplied URLs are passed directly to libcurl without validation of internal IP ranges, redirects are automatically followed, and there is no validation of redirect destinations at the Python layer. An attacker can provide a URL pointing to their server, which redirects to an internal service, allowing access to internal endpoints. The TLS fingerprint impersonation feature can further help bypass TLS-based filtering controls. Recommendations 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** GMT versions 6.6.0 and prior **Description** GMT is an open source collection of command-line tools for manipulating geographic and Cartesian data sets. A stack-based buffer overflow exists in the `gmt remote dataset id` function within `src/gmt remote.c`. This occurs when a specially crafted long string is provided as a dataset identifier, potentially leading to a crash or arbitrary code execution. The issue is triggered when using the which module. This has been addressed with commit 0ad2b49. **Recommendations** Versions prior to 6.6.0 should be updated to a newer version that includes commit 0ad2b49.

**Name of the Vulnerable Software and Affected Versions** Intake versions prior to 2.0.9 **Description** Intake is a package used for finding, investigating, loading, and disseminating data. A flaw exists where the shell() syntax within parameter default values is automatically expanded during catalog parsing. If a catalog contains a parameter default such as shell(<command>), the command may be executed when the catalog source is accessed. This allows an attacker to execute commands on the host system by embedding them within a malicious catalog YAML file. The issue originates from how parameter default values are expanded during catalog loading and source access, specifically involving the `expand defaults()` function and related parameter parsing mechanisms. A proof-of-concept demonstrates that a malicious YAML file can be crafted to execute arbitrary commands, such as creating a file in the /tmp directory. An attacker could distribute a malicious catalog file through various channels, including Git repositories, shared datasets, and URLs. A successful attack could allow an attacker to execute commands with the local user privileges of the victim. **Recommendations** Versions prior to 2.0.9 should be updated to version 2.0.9 or later.

**Name of the Vulnerable Software and Affected Versions** dynaconf versions prior to 3.2.13 **Description** dynaconf is susceptible to Server-Side Template Injection (SSTI) due to insecure template evaluation within the @Jinja resolver. When the `jinja2` package is installed, Dynaconf evaluates template expressions embedded in configuration values without a sandboxed environment. Attackers influencing configuration sources—such as environment variables, `.env` files, container environment configuration, or CI/CD secrets—can execute arbitrary OS commands on the host system. The @format resolver also allows object graph traversal, potentially exposing sensitive runtime objects and environment variables. A proof-of-concept (PoC) demonstrates command execution using the `cycler` object to access Python’s `os` module and execute commands. Successful exploitation can lead to arbitrary OS command execution, access to sensitive environment variables, compromise of application secrets, and full compromise of the running application process. The vulnerability can be remotely exploitable in real-world deployments due to the potential for configuration values to originate from CI/CD pipelines, container orchestration systems, or environment injection. **Recommendations** versions prior to 3.2.13: Upgrade to version 3.2.13 or later to address the vulnerability. versions prior to 3.2.13: Use Jinja2 sandbox for template rendering. versions prior to 3.2.13: Restrict @format usage to trusted values.

**Name of the Vulnerable Software and Affected Versions** eml parser versions prior to 2.0.1 **Description** The eml parser module, used for parsing eml files, contains a path traversal issue in the example script `examples/recursively extract attachments.py`. This allows for arbitrary file write outside the intended output directory. Attachment filenames extracted from parsed emails are directly used to construct output file paths without sanitization. An attacker-controlled filename can escape the target directory by crafting email attachment headers, such as setting the `Content-Disposition` header with a filename like `../outside/pwned.txt`. The vulnerable code is located in lines 61-64 of the `examples/recursively extract attachments.py` script, where the `a['filename']` variable is used directly to construct the output file path. The vulnerability is limited to the example script and does not affect the core eml parser library. Potential attack scenarios include cron job injection, web shell upload, and SSH key injection. The vulnerable parameter is `a['filename']`. **Recommendations** Versions prior to 2.0.1 should be updated to version 2.0.1 or later. As a temporary workaround, implement path normalization and boundary validation before writing files. Specifically, use `os.path.basename()` to extract the filename and ensure the resulting path is relative to the intended output directory using `out filepath.resolve().is relative to(out path.resolve())`.