| CVE |
Vendors |
Products |
Updated |
CVSS v3.1 |
| Heap-based buffer overflow in Windows Libarchive allows an unauthorized attacker to execute code over a network. |
| Untrusted pointer dereference in Windows Distributed File System (DFS) allows an unauthorized attacker to elevate privileges over a network. |
| vLLM is an inference and serving engine for large language models. From 0.16.0 until 0.32.0, the /v1/responses endpoint allows an authenticated caller to place an attacker-controlled template in chat_template_kwargs[chat_template] when no server template is configured. ChatParams.get_apply_chat_template_kwargs() preserves that value without enforcing trust_request_chat_template, so the request can bypass the intended trust policy and cause CPU or memory exhaustion through Jinja rendering. This issue is fixed in version 0.32.0. |
| Use after free in Windows Broadcast DVR User Service allows an authorized attacker to elevate privileges locally. |
| vLLM is an inference and serving engine for large language models. Prior to 0.32.0, the OpenAI-compatible server configures Starlette CORSMiddleware with allowed_origins set to a wildcard by default. A malicious web page can use a victim's browser to reach an unauthenticated vLLM server on localhost, an internal network, or a VPN, call endpoints including GET /v1/models and POST /v1/chat/completions, consume GPU resources, and read responses. Cookie-based cross-site request forgery does not apply because credentials are disabled by default, and deployments requiring bearer authentication are exposed only when the caller already has a usable token. This issue is fixed in version 0.32.0. |
| nforwardauth is a forward auth service that allows the use of a single authentication middleware for multiple sites. Versions prior to 1.5.1 accept an unvalidated `r` query parameter on the `/login` endpoint and use them as the post-authentication redirect destination. An unauthenticated attacker can create a legitimate `nforwardauth` login URL that redirects a victim to an attacker-controlled website after the victim successfully authenticates. This issue may facilitate phishing and other social-engineering attacks by abusing the trust associated with the legitimate authentication domain. This vulnerability does not provide an unauthenticated ForwardAuth bypass. Version 1.5.1 contains a patch. |
| vLLM is an inference and serving engine for large language models. Prior to 0.32.0, vllm/tokenizers/hf.py computes max_token_id with an off-by-one upper bound, allowing a pre-tokenized prompt ID equal to the vocabulary size to pass validation. For a tensor-parallel size of one and an exact multiple-of-64 vocabulary, VocabParallelEmbedding.forward does not mask that value before F.embedding uses it, causing a CUDA device-side assertion and terminating the engine. This issue is fixed in version 0.32.0. |
| An issue in libass v.0.17.5 allows a remote attacker to cause a denial of service via the ass_outline_construct() in libass/ass_render.c |
| IBM Carbon Charts 0.4.0 through 1.27.19 is vulnerable to cross-site scripting. The MeterChart component's Title.truncateTitle() function re-inserts a substring of user-supplied label text into the DOM via an unsanitized innerHTML write when the title exceeds the available container width. A remote attacker could exploit this vulnerability to execute arbitrary JavaScript in the victim's browser session. |
| Path Traversal and Missing Authorization in cloud-storage source in Google Cloud MCP Toolbox for Databases (googleapis/mcp-toolbox) versions 1.2.0 through 1.4.0 allows an unauthenticated remote attacker to exfiltrate host files to external Google Cloud Storage buckets and overwrite arbitrary local files via crafted tool invocation parameters. |
| An issue in SKYWORTH DIGITAL CO., LTD SK-WR6640X Wi-Fi 6 Dual-Band Fiber Router (AX1800M) V1.0.3 allows a remote attacker to obtain sensitive information via the hidden_factory_switch.gch |
| boneIO Black is vulnerable to Cross-Site Request Forgery (CSRF) on all state-changing API endpoints because it lacks CSRF tokens and does not validate Origin or Referer headers. A remote attacker can craft a malicious webpage that, when visited by a user on the internal network, sends cross-origin requests to the device. This allows the attacker to reboot the controller, toggle relay circuits, control gates and blinds, or shut down the device, causing loss of integrity and availability. Critically, due to CVE-2026-77053 (authentication disabled by default), any user on the same internal network can trigger attacker-controlled actions by visiting the malicious website.
This issue was fixed in version 1.6.0. |
| boneIO black_debian_images ships with a Mosquitto MQTT broker that uses the same default credentials on all devices. These credentials are publicly documented on github, and there is no indication in the instructions to change them. A remote attacker can use them to gain full read and write access to the broker, allowing control of connected relays, blinds, and gates, and injection of fake entities into a linked Home Assistant installation.
This issue was fixed in version 1.6.0 of the image: MQTT passwords are generated per device at first boot. Devices upgraded from earlier versions keep their existing broker credentials; from boneio 1.6.0 the web panel reports credentials that are still the published defaults, and the owner has to change them. |
| boneIO Black stores passwords, including the MQTT broker and web panel passwords, in cleartext. The GET /api/config endpoint returns the full configuration with sub-services passwords exposed. Any authenticated user can call the endpoint and access the administrative passwords in cleartext form. Critically, due to CVE-2026-77053 (authentication disabled by default), an unauthenticated attacker on the same local network can retrieve all passwords.
This issue was fixed in version 1.6.0: web panel passwords are stored as salted scrypt hashes in a file readable only by the service account, and the API and logs no longer return secret values. The MQTT broker password has to stay recoverable because the application presents it to the broker as a client; it is kept in a separate file readable only by the service account and is no longer exposed through the API. |
| In Splunk SOAR versions below 8.7.0, an authenticated user could change the Enterprise Security edition through the Representational State Transfer (REST) API and cause Splunk SOAR to apply the license change.
The vulnerability is possible because Splunk SOAR authenticates requests to manage the Enterprise Security edition but does not authorize the caller before it applies the license change.
For more information see Using the REST API reference for Splunk SOAR (On-premises) (https://help.splunk.com/en/splunk-soar/soar-on-premises/rest-api-reference/8.7.0/using-the-splunk-soar-rest-api/using-the-rest-api-reference-for-splunk-soar-on-premises), Obtain and configure a Splunk SOAR (On-premises) license (https://help.splunk.com/en/splunk-soar/soar-on-premises/administer-soar-on-premises/8.7.0/configure-your-companys-settings-in-splunk-soar-on-premises/obtain-and-configure-a-splunk-soar-on-premises-license), and Integration of Splunk SOAR with Splunk Enterprise Security (https://help.splunk.com/en/splunk-enterprise-security-8/administer/8.7/automation-with-playbooks/integration-of-splunk-soar-with-splunk-enterprise-security) in the Splunk documentation. |
| In Splunk SOAR versions below 8.7.0, an authenticated user who does not hold a role with permission to edit Automation Brokers could download the Splunk SOAR Automation Broker Docker Compose configuration file. The file contains the Uniform Resource Locator (URL) of the Splunk SOAR instance, the container image location, and the Splunk SOAR Automation Broker version.
The vulnerability is possible because Splunk SOAR returns the configuration file without enforcing the Automation Broker edit permission.
For more information see Install Splunk Automation Broker on a Docker host (https://help.splunk.com/en/splunk-soar/splunk-automation-broker/install-and-create-classic-splunk-soar-automation-brokers/install-splunk-automation-broker-on-a-docker-host) and Manage roles and permissions in Splunk SOAR (On-premises) (https://help.splunk.com/en/splunk-soar/soar-on-premises/administer-soar-on-premises/8.5.0/manage-your-splunk-soar-on-premises-users-and-accounts/manage-roles-and-permissions-in-splunk-soar-on-premises) in the Splunk documentation. |
| In Splunk Enterprise versions below 10.4.4, 10.2.8, 10.0.11, and 9.4.16, a user who does not hold the "admin" or "power" Splunk roles could create, move, edit, or delete sourcetype renaming configurations that they own in apps that they do not have permission to access.
The vulnerability is possible when Splunk Web is turned on because Splunk Web does not verify the required app permissions before it performs sourcetype renaming operations.
For more information see Rename source types at search time (https://help.splunk.com/en/splunk-enterprise/get-started/get-data-in/10.6/configure-source-types/rename-source-types-at-search-time) in the Splunk documentation. |
| The root-privileged `wsserver` terminal-bridge daemon builds a shell command from client-supplied WebSocket fields and passes it to `system()` and `popen()` without quoting, escaping, or character-set validation. The daemon's public-key session path performs no appliance-session check. |
| Due to improper permission checks in Lobster_pro, attackers are able to interact with endpoints that should require administrative privileges. This issue affects Lobster_pro version 4.12.12-LTS and potentially other versions prior to 25.1.3. |
| Open5GS 2.6.1 through 2.8.0 contains an out-of-bounds read vulnerability in parse_scheme_output() in lib/sbi/conv.c during SUCI de-concealment due to incomplete length validation. Remote unauthenticated UEs or SBI peers can submit a crafted SUCI with a 42-byte scheme output to a UDM, triggering a heap over-read past the buffer. |
