A path traversal vulnerability in GitPython allows attackers who can supply crafted reference paths to write, overwrite, move, or delete files outside the repository's .git directory. This occurs due to insufficient validation of reference paths in creation, rename, and delete operations. The vulnerability affects GitPython versions up to 3.1.46 and the current main branch (3.1.47 locally). It can lead to high integrity and availability impacts by corrupting or deleting files accessible to the process. A fix is available in version 3.0.4.
GitPython versions from 2.9.2 up to but not including 3.0.4 contain a vulnerability in the handling of multi_options in the clone and submodule update functions. The vulnerability arises because validation is performed on the original list of options before they are split by shlex.split, allowing an attacker to embed unsafe options like '--config core.hooksPath=...' within a single string that passes validation but is split into multiple unsafe tokens at execution time. This can lead to execution of attacker-controlled Git hooks during repository cloning.
GitPython versions from 2.9.2 up to but not including 3.0.4 contain a command injection vulnerability due to improper validation of Python keyword arguments that map to unsafe Git command options. Specifically, the use of underscore-form kwargs like 'upload_pack' and 'receive_pack' bypasses the unsafe option checks and allows arbitrary command execution when passed to methods such as Repo.clone_from(), Remote.fetch(), Remote.pull(), or Remote.push(). This vulnerability is addressed in version 3.0.4.
CVE-2026-41066 is a vulnerability in the lxml library where the default configuration of the iterparse() and ETCompatXMLParser() functions allows XML External Entity (XXE) attacks that can read local files. This occurs because the default setting for resolve_entities is true, enabling entity resolution that can access local files. The issue affects lxml versions from 2.9.2 up to but not including 3.0.4. The vulnerability is fixed in version 3.0.4, which changes the default to resolve_entities='internal', preventing local file access by default.
CairoSVG, used by Standardebooks, is vulnerable to an exponential denial of service (DoS) via recursive <use> element amplification. The vulnerability arises from the use() function in cairosvg/defs.py recursively processing <use> elements without limits, causing CPU exhaustion from small SVG inputs. This can cause services that render SVGs to hang indefinitely at 100% CPU usage without crashing. A fix is available in version 3.0.3 that adds recursion depth limits.
A Regular Expression Denial of Service (ReDoS) vulnerability exists in soupsieve's CSS selector parser due to a regex pattern that suffers catastrophic backtracking when processing attribute selectors with unterminated quoted values. This can cause CPU exhaustion and thread blocking with a payload as small as 300 bytes. The issue affects applications that compile untrusted CSS selectors using soupsieve or Beautiful Soup's select methods. A fix is available in version 2.12.1_2.
The soupsieve CSS selector parser used by Beautiful Soup 4 allocates unbounded memory when processing large comma-separated selector lists. An attacker can supply a crafted CSS selector string with many comma-separated items, causing the parser to allocate hundreds of megabytes of heap memory from a relatively small input, leading to memory exhaustion and denial of service. This affects applications that use soupsieve.compile() or Beautiful Soup's .select()/.select_one() with user-supplied selectors. The issue is fixed in version 2.12.1_2.
A vulnerability in ansible-core's ansible-galaxy role install command allows a malicious role author to inject arbitrary git configuration flags via the src field in a role's meta/requirements.yml file. This improper neutralization of argument delimiters can lead to arbitrary code execution on the machine of a user who installs the malicious role. The flaw affects ansible-core versions from 4.0.0 up to but not including 14.0.0_2. A fix is available in version 14.0.0_2. Exploitation requires user interaction by installing a crafted role, but the impact is severe, potentially leading to full system compromise.
Ansible Community General Collection versions prior to 14.3.0 contain a vulnerability that exposes sensitive credentials, including plaintext passwords, in verbose debug output. This flaw allows attackers with access to logs or debug output to retrieve these secrets, potentially compromising Keycloak accounts or administrative access. The issue arises from the lack of proper output suppression for password-containing data structures. A fix is available in version 14.3.0.
A vulnerability in ansible-core's user module allows an unprivileged user to create or overwrite files on any system path and take ownership when a privileged user runs the module against the unprivileged user's home directory. This can lead to unauthorized read/write access to files if directory traversal permissions exist. The issue affects ansible versions prior to 14.3.0 and has been assigned CVE-2024-9902 with a CVSS score of 6.3 (medium severity).
CVE-2024-8775 is a vulnerability in Ansible-core where sensitive information stored in Ansible Vault files can be exposed in plaintext during playbook execution if the no_log: true parameter is not set when loading vaulted variables. This can lead to unintentional disclosure of secrets such as passwords or API keys in logs or output. The issue affects versions prior to 14.3.0 and has been fixed in version 14.3.0.
Ansible-Core contains a vulnerability (CVE-2024-11079) that allows attackers to bypass unsafe content protections by leveraging the hostvars object to reference and execute templated content. This can lead to arbitrary code execution if remote data or module outputs are improperly templated within playbooks. The issue affects versions prior to 14.3.0 and has been assigned a medium severity with a CVSS score of 5.5. A fix is available in version 14.3.0.
Ansible-core versions prior to 14.3.0 contain an information disclosure vulnerability where certain tasks do not respect the ANSIBLE_NO_LOG configuration, potentially exposing sensitive data such as decrypted secrets in output, including loop items. This flaw has been assigned CVE-2024-0690 and has a medium severity rating with a CVSS score of 5. A fix is available in version 14.3.0.
CVE-2026-22701 is a Time-of-Check-Time-of-Use (TOCTOU) race condition vulnerability in the SoftFileLock class of the filelock package used by ansible-lint. It allows a local attacker with permission to create symlinks to exploit a race window between permission checking and file creation, potentially causing lock operations to fail, leading to denial of service or unintended file access. The vulnerability affects applications using SoftFileLock or fallback FileLock on systems without fcntl support. A patch adding the O_NOFOLLOW flag to prevent symlink following during lock file creation is available in version 26.1.1. Users are advised to upgrade or apply mitigations such as restricting filesystem permissions and avoiding SoftFileLock in sensitive contexts.
A TOCTOU (Time-of-Check-Time-of-Use) race condition in the filelock library allows local attackers to exploit a window between checking for a lock file's existence and opening it with truncation flags. By creating a symlink during this window, an attacker can cause the victim application to truncate arbitrary files, potentially corrupting or deleting critical data. This affects Unix, Linux, macOS, and Windows systems using filelock, impacting dependent tools like virtualenv, PyTorch, poetry, and tox. The vulnerability requires local filesystem access and the ability to create symlinks. It is fixed in filelock version 3.20.1 by adding safeguards against symlink following during lock acquisition.