ghsa

### Summary If untrusted user input is used to dynamically create a `PebbleTemplate` with the method `PebbleEngine#getLiteralTemplate`, then an attacker can include arbitrary local files from the file system into the generated template, leaking potentially sensitive information into the output of `PebbleTemplate#evaluate`. This is done via the `include` macro. ### Details The `include` macro calls `PebbleTempateImpl#resolveRelativePath` with the `relativePath` argument passed within the template: Example template: ``` {% include [relativePath] %} ``` When `resolveRelativePath` is called, the `relativePath` is resolved against the `PebbleTemplateImpl.name` variable. ```java /** * This method resolves the given relative path based on this template file path. * * @param relativePath the path which should be resolved. * @return the resolved path. */ public String resolveRelativePath(String relativePath) { String resolved = this.engine.getLoader().resolveRelativ...

Two denial of service vulnerabilities were found in ntpd-rs related to the handling of NTS cookies in our client functionality. Whenever an NTS source is configured and the server behind that source is sending zero-sized cookies or cookies larger than what would fit in our buffer size, ntpd-rs would crash. Only configured NTS sources can abuse these vulnerabilities. NTP sources or third parties that are not configured cannot make use of these vulnerabilities. For zero-sized cookies: a division by zero would force an exit when the number of new cookies that would need to be requested is calculated. In ntpd-rs 1.5.0 a check was added to prevent the division by zero. For large cookies: while trying to send a NTP request with the cookie included, the buffer is too small to handle the cookie and an exit of ntpd-rs is forced once a write to the buffer is attempted. The memory outside the buffer would not be written to in this case. In ntpd-rs 1.5.0 a check was added that prevents accepting...

Name: ASA-2025-004: Non-deterministic JSON Unmarshalling of IBC Acknowledgement can result in a chain halt Component: IBC-Go Criticality: Critical (Considerable Impact; Almost Certain Likelihood per [ACMv1.2](https://github.com/interchainio/security/blob/main/resources/CLASSIFICATION_MATRIX.md)) Affected versions: IBC-Go >= v7; Earlier IBC-Go versions may also be affected. Affected users: Validators, Full nodes, IBC Middleware authors ### Description An issue was discovered in IBC-Go's deserialization of acknowledgements that results in non-deterministic behavior which can halt a chain. Any user that can open an IBC channel can introduce this state to the chain ### Patches The new IBC-Go releases below address this issue: * [v7.9.2](https://github.com/cosmos/ibc-go/releases/tag/v7.9.2) * [v8.6.1](https://github.com/cosmos/ibc-go/releases/tag/v8.6.1) ### Workarounds To prevent this state from being introduced to a chain, it is possible to permission Channel Opening as a workaroun...

### Summary The `CacheHandler` class creates a cache file to store the auth token here: https://github.com/spotipy-dev/spotipy/blob/master/spotipy/cache_handler.py#L93-L98 The file created has `rw-r--r--` (644) permissions by default, when it could be locked down to `rw-------` (600) permissions. I think `600` is a sensible default.  ### Details This leads to overly broad exposure of the spotify auth token. If this token can be read by an attacker (another user on the machine, or a process running as another user), it can be used to perform administrative actions on the Spotify account, depending on the scope granted to the token. ### PoC Run an application that uses spotipy with client creation like this: ```python from pathlib import Path import spotipy from os import getenv def create_spotify_client(client_id: str, client_secret: str) -> spotipy.Spotify: """Create and return an auth...

elestio memos v0.23.0 is vulnerable to Server-Side Request Forgery (SSRF) due to insufficient validation of user-supplied URLs, which can be exploited to perform SSRF attacks.

mongosh may be susceptible to local privilege escalation under certain conditions potentially enabling unauthorized actions on a user's system with elevated privilege, when a crafted file is stored in C:\node_modules\. This issue affects mongosh prior to 2.3.0.

### Impact A vulnerability in Rancher has been discovered, leading to a local user impersonation through SAML Authentication on first login. The issue occurs when a SAML authentication provider (AP) is configured (e.g. Keycloak). A newly created AP user can impersonate any user on Rancher by manipulating cookie values during their initial login to Rancher. This vulnerability could also be exploited if a Rancher user (present on the AP) is removed, either manually or automatically via the [User Retention feature](https://ranchermanager.docs.rancher.com/how-to-guides/advanced-user-guides/enable-user-retention) with delete-inactive-user-after. More precisely, Rancher validates only a subset of input from the SAML assertion request; however, it trusts and uses values that are not properly validated. An attacker could then configure the saml_Rancher_UserID cookie and the saml_Rancher_Action cookie so that the user principal from the AP will be added to the user specified by the attacker (...

### Impact An unauthenticated stack overflow crash, leading to a denial of service (DoS), was identified in Rancher’s `/v3-public/authproviders` public API endpoint. A malicious user could submit data to the API which would cause the Rancher server to crash, but no malicious or incorrect data would actually be written in the API. The downstream clusters, i.e., the clusters managed by Rancher, are not affected by this issue. This vulnerability affects those using external authentication providers as well as Rancher’s local authentication. ### Patches The patch includes the removal of unnecessary HTTP methods of the specific API. Patched versions include releases `v2.8.13`, `v2.9.7` and `v2.10.3`. ### Workarounds There are no workarounds for this issue. Users are recommended to upgrade, as soon as possible, to a version of Rancher Manager that contains the fix. ### References If you have any questions or comments about this advisory: - Reach out to the [SUSE Rancher Security team](h...

### Impact A vulnerability has been identified within Rancher where it is possible for an unauthenticated user to list all CLI authentication tokens and delete them before the CLI is able to get the token value. This effectively prevents users from logging in via the CLI when using rancher token as the execution command (instead of the token directly being in the kubeconfig). Note that this token is not the kubeconfig token and if an attacker is able to intercept it they can't use it to impersonate a real user since it is encrypted. This happens because for SAML-based authentication providers, the login flow from the CLI works by generating a link to be pasted in the browser, and then polling every 10 seconds for the `/v3-public/authTokens/<token name>` endpoint. The `<token name>` is randomly generated by the CLI. Once the login flow succeeds, Rancher creates an auth token (with an encrypted token value). The CLI then deletes the authToken. Rancher deployments using only the loc...

The MongoDB Shell may be susceptible to control character injection where an attacker with control over the database cluster contents can inject control characters into the shell output. This may result in the display of falsified messages that appear to originate from mongosh or the underlying operating system, potentially misleading users into executing unsafe actions. The vulnerability is exploitable only when mongosh is connected to a cluster that is partially or fully controlled by an attacker. This issue affects mongosh versions prior to 2.3.9.