CVE’s

Tensorflow is an Open Source Machine Learning Framework. An attacker can craft a TFLite model that would cause an integer overflow in embedding lookup operations. Both `embedding_size` and `lookup_size` are products of values provided by the user. Hence, a malicious user could trigger overflows in the multiplication. In certain scenarios, this can then result in heap OOB read/write. Users are advised to upgrade to a patched version.

Tensorflow is an Open Source Machine Learning Framework. An attacker can craft a TFLite model that would cause an integer overflow in `TfLiteIntArrayCreate`. The `TfLiteIntArrayGetSizeInBytes` returns an `int` instead of a `size_t. An attacker can control model inputs such that `computed_size` overflows the size of `int` datatype. The fix will be included in TensorFlow 2.8.0. We will also cherrypick this commit on TensorFlow 2.7.1, TensorFlow 2.6.3, and TensorFlow 2.5.3, as these are also affected and still in supported range.

Tensorflow is an Open Source Machine Learning Framework. An attacker can craft a TFLite model that would trigger a division by zero in `BiasAndClamp` implementation. There is no check that the `bias_size` is non zero. The fix will be included in TensorFlow 2.8.0. We will also cherrypick this commit on TensorFlow 2.7.1, TensorFlow 2.6.3, and TensorFlow 2.5.3, as these are also affected and still in supported range.

CodeIgniter is a PHP full-stack web framework. This vulnerability may allow attackers to spoof their IP address when the server is behind a reverse proxy. This issue has been patched, please upgrade to version 4.2.11 or later, and configure `ConfigApp::$proxyIPs`. As a workaround, do not use `$request->getIPAddress()`.

authentik is an open-source Identity Provider focused on flexibility and versatility. Versions prior to 2022.11.4 and 2022.10.4 are vulnerable to Improper Authentication. Token reuse in invitation URLs leads to access control bypass via the use of a different enrollment flow than in the one provided. The vulnerability allows an attacker that knows different invitation flows names (e.g. `enrollment-invitation-test` and `enrollment-invitation-admin`) via either different invite links or via brute forcing to signup via a single invitation url for any valid invite link received (it can even be a url for a third flow as long as it's a valid invite) as the token used in the `Invitations` section of the Admin interface does NOT change when a different `enrollment flow` is selected via the interface and it is NOT bound to the selected flow, so it will be valid for any flow when used. This issue is patched in authentik 2022.11.4,2022.10.4 and 2022.12.0. Only configurations that use invitations and have multiple enrollment flows with invitation stages that grant different permissions are affected. The default configuration is not vulnerable, and neither are configurations with a single enrollment flow. As a workaround, fixed data can be added to invitations which can be checked in the flow to deny requests. Alternatively, an identifier with high entropy (like a UUID) can be used as flow slug, mitigating the attack vector by exponentially decreasing the possibility of discovering other flows.

Alpine is a scaffolding library in Java. Alpine prior to version 1.10.4 allows Authentication Filter bypass. The AuthenticationFilter relies on the request URI to evaluate if the user is accessing the swagger endpoint. By accessing a URL with a path such as /api/foo;%2fapi%2fswagger the contains condition will hold and will return from the authentication filter without aborting the request. Note that the principal object will not be assigned and therefore the issue wont allow user impersonation. This issue has been fixed in version 1.10.4. There are no known workarounds.

Alpine is a scaffolding library in Java. Alpine prior to version 1.10.4 allows URL access filter bypass. This issue has been fixed in version 1.10.4. There are no known workarounds.

Grafana is an open-source platform for monitoring and observability. Starting with the 8.1 branch and prior to versions 8.5.16, 9.2.10, and 9.3.4, Grafana had a stored XSS vulnerability affecting the core plugin GeoMap. The stored XSS vulnerability was possible because SVG files weren't properly sanitized and allowed arbitrary JavaScript to be executed in the context of the currently authorized user of the Grafana instance. An attacker needs to have the Editor role in order to change a panel to include either an external URL to a SVG-file containing JavaScript, or use the `data:` scheme to load an inline SVG-file containing JavaScript. This means that vertical privilege escalation is possible, where a user with Editor role can change to a known password for a user having Admin role if the user with Admin role executes malicious JavaScript viewing a dashboard. Users may upgrade to version 8.5.16, 9.2.10, or 9.3.4 to receive a fix.

aad-pod-identity assigns Azure Active Directory identities to Kubernetes applications and has now been deprecated as of 24 October 2022. The NMI component in AAD Pod Identity intercepts and validates token requests based on regex. In this case, a token request made with backslash in the request (example: `/metadata/identityoauth2token/`) would bypass the NMI validation and be sent to IMDS allowing a pod in the cluster to access identities that it shouldn't have access to. This issue has been fixed and has been included in AAD Pod Identity release version 1.8.13. If using the AKS pod-managed identities add-on, no action is required. The clusters should now be running the version 1.8.13 release.

The Easy Username Updater WordPress plugin before 1.0.5 does not implement CSRF checks, which could allow attackers to make a logged in admin change any user's username includes the admin