A Cross-site Scripting (XSS) vulnerability exists in the Settings page of parisneo/lollms-webui version 9.8. The vulnerability is due to the improper use of the 'v-html' directive, which inserts the content of the 'full_template' variable directly as HTML. This allows an attacker to execute malicio…

SuperAGI version v0.0.14 is vulnerable to an unauthenticated Denial of Service (DoS) attack. The vulnerability exists in the resource upload request, where appending characters, such as dashes (-), to the end of a multipart boundary in an HTTP request causes the server to continuously process each …

An improper access control vulnerability in open-webui/open-webui v0.3.8 allows attackers to view and delete any files. The application does not verify whether the attacker is an administrator, allowing the attacker to directly call the GET /api/v1/files/ interface to retrieve information on all fi…

vllm-project vllm version v0.6.2 contains a vulnerability in the MessageQueue.dequeue() API function. The function uses pickle.loads to parse received sockets directly, leading to a remote code execution vulnerability. An attacker can exploit this by sending a malicious payload to the MessageQueue,…

In lunary-ai/lunary version v1.4.29, the GET /projects API endpoint exposes both public and private API keys for all projects to users with minimal permissions, such as Viewers or Prompt Editors. This vulnerability allows unauthorized users to retrieve sensitive credentials, which can be used to pe…

In lunary-ai/lunary, the privilege check mechanism is flawed in version git afc5df4. The system incorrectly identifies certain endpoints as public if the path contains '/auth/' anywhere within it. This allows unauthenticated attackers to access sensitive endpoints by including '/auth/' in the path.…

In the latest version of pytorch/serve, the script 'upload_results_to_s3.sh' references the S3 bucket 'benchmarkai-metrics-prod' without ensuring its ownership or confirming its accessibility. This could lead to potential security vulnerabilities or unauthorized access to the bucket if it is not pr…

In binary-husky/gpt_academic version <= 3.83, the plugin `CodeInterpreter` is vulnerable to code injection caused by prompt injection. The root cause is the execution of user-provided prompts that generate untrusted code without a sandbox, allowing the execution of parts of the LLM-generated code. …

An open redirect vulnerability in haotian-liu/llava version v1.2.0 (LLaVA-1.6) allows a remote unauthenticated attacker to redirect users to arbitrary websites via a specially crafted URL. This can be exploited for phishing attacks, malware distribution, and credential theft.

A vulnerability in binary-husky/gpt_academic version git 310122f allows for remote code execution. The application supports the extraction of user-provided RAR files without proper validation. The Python rarfile module, which supports symlinks, can be exploited to perform arbitrary file writes. Thi…