#git

The huggingface/transformers library is vulnerable to arbitrary code execution through deserialization of untrusted data within the `load_repo_checkpoint()` function of the `TFPreTrainedModel()` class. Attackers can execute arbitrary code and commands by crafting a malicious serialized payload, exploiting the use of `pickle.load()` on data from potentially untrusted sources. This vulnerability allows for remote code execution (RCE) by deceiving victims into loading a seemingly harmless checkpoint during a normal training process, thereby enabling attackers to execute arbitrary code on the targeted machine.

A critical Remote Code Execution (RCE) vulnerability was identified in the aimhubio/aim project, specifically within the `/api/runs/search/run/` endpoint, affecting versions >= 3.0.0. The vulnerability resides in the `run_search_api` function of the `aim/web/api/runs/views.py` file, where improper restriction of user access to the `RunView` object allows for the execution of arbitrary code via the `query` parameter. This issue enables attackers to execute arbitrary commands on the server, potentially leading to full system compromise.

A vulnerability was identified in the `exec_utils` class of the `llama_index` package, specifically within the `safe_eval` function, allowing for prompt injection leading to arbitrary code execution. This issue arises due to insufficient validation of input, which can be exploited to bypass method restrictions and execute unauthorized code. The vulnerability is a bypass of the previously addressed CVE-2023-39662, demonstrated through a proof of concept that creates a file on the system by exploiting the flaw.

aimhubio/aim is vulnerable to Cross-Site Request Forgery (CSRF), allowing attackers to perform actions such as deleting runs, updating data, and stealing data like log records and notes without the user's consent. The vulnerability stems from the lack of CSRF and CORS protection in the aim dashboard. An attacker can exploit this by tricking a user into executing a malicious script that sends unauthorized requests to the aim server, leading to potential data loss and unauthorized data manipulation.

gradio-app/gradio is vulnerable to a local file inclusion vulnerability due to improper validation of user-supplied input in the UploadButton component. Attackers can exploit this vulnerability to read arbitrary files on the filesystem, such as private SSH keys, by manipulating the file path in the request to the `/queue/join` endpoint. This issue could potentially lead to remote code execution. The vulnerability is present in the handling of file upload paths, allowing attackers to redirect file uploads to unintended locations on the server.

A command injection vulnerability exists in the `TranscriptEndpoint` of mudler/localai, specifically within the `audioToWav` function used for converting audio files to WAV format for transcription. The vulnerability arises due to the lack of sanitization of user-supplied filenames before passing them to ffmpeg via a shell command, allowing an attacker to execute arbitrary commands on the host system. Successful exploitation could lead to unauthorized access, data breaches, or other detrimental impacts, depending on the privileges of the process executing the code.

BerriAI/litellm is vulnerable to Server-Side Template Injection (SSTI) via the `/completions` endpoint. The vulnerability arises from the `hf_chat_template` method processing the `chat_template` parameter from the `tokenizer_config.json` file through the Jinja template engine without proper sanitization. Attackers can exploit this by crafting malicious `tokenizer_config.json` files that execute arbitrary code on the server.

### Impact Parameters of UI extensions are always interpreted as Velocity code and executed with programming rights. Any user with edit right on any document like the user's own profile can create UI extensions. This allows remote code execution and thereby impacts the confidentiality, integrity and availability of the whole XWiki installation. To reproduce, edit your user profile with the object editor and add a UIExtension object with the following values: ``` Extension Point ID: org.xwiki.platform.panels.Applications Extension ID: platform.panels.myFakeApplication Extension parameters:  label=I got programming right: $services.security.authorization.hasAccess('programming') target=Main.WebHome targetQueryString= icon=icon:bomb Extension Scope: "Current User". ``` Save the document and open any document. If an application entry with the text "I got programming right: true" is displayed, the attack succeeded, if the code in "label" is displayed literally, the XWiki installation isn'...

### Impact The HTML escaping of escaping tool that is used in XWiki doesn't escape `{`, which, when used in certain places, allows XWiki syntax injection and thereby remote code execution. To reproduce in an XWiki installation, open `<xwiki-host>/xwiki/bin/view/Panels/PanelLayoutUpdate?place=%7B%7B%2Fhtml%7D%7D%7B%7Basync%20async%3Dfalse%7D%7D%7B%7Bvelocity%7D%7D%23evaluate(%24request.eval)%7B%7B%2Fvelocity%7D%7D%7B%7B%2Fasync%7D%7D&eval=Hello%20from%20URL%20Parameter!%20I%20got%20programming%3A%20%24services.security.authorization.hasAccess(%27programming%27)` where `<xwiki-host>` is the URL of your XWiki installation. If this displays `You are not admin on this place Hello from URL Parameter! I got programming: true`, the installation is vulnerable. ### Patches The vulnerability has been fixed on XWiki 14.10.19, 15.5.5, and 15.9 RC1. ### Workarounds Apart from upgrading, there is no generic workaround. However, replacing `$escapetool.html` by `$escapetool.xml` in XWiki documents f...

### Impact When invoking a capability with a chain depth of 2, i.e., it is delegated directly from the root capability, the `expires` property is not properly checked against the current date or other `date` param. This can allow invocations outside of the original intended time period. A zcap still cannot be invoked without being able to use the associated private key material. ### Patches `@digitalbazaar/zcap` v9.0.1 fixes expiration checking. ### Workarounds A zcap could be revoked at any time. ### References https://github.com/digitalbazaar/zcap/pull/82