Improper restriction of intended endpoints in the RabbitMQ broker connection tools of the Amazon MQ MCP Server (awslabs.amazon-mq-mcp-server) before 2.0.24 may allow a remote unauthenticated actor (via prompt injection) to obtain Amazon MQ for RabbitMQ broker credentials or OAuth access tokens sent to a crafted endpoint controlled through a broker hostname introduced in the MCP client context. To remediate this issue, users should upgrade to version 2.0.24.

A stored cross-site scripting (XSS) vulnerability exists in osTicket 1.18.3 due to improper sanitization of the thread entry title field. User-controlled input in the title is stored without adequate HTML escaping and later rendered in multiple staff-facing templates without proper output encoding. An attacker can inject arbitrary JavaScript by submitting a crafted ticket reply or email with a malicious subject line.

WebsiteBaker CMS before 2.13.10 contains an unrestricted file upload vulnerability in the module installation feature that allows authenticated administrators to achieve remote code execution by uploading a crafted ZIP archive containing a PHP webshell alongside a valid info.php metadata file. Attackers can place the malicious archive through the module installation interface, causing the application to extract the webshell into a web-accessible modules/ subdirectory where it becomes immediately executable by any unauthenticated user via direct HTTP request.

Krayin CRM 2.2.4 contains a missing authentication vulnerability in the installer middleware that allows unauthenticated remote attackers to overwrite the primary administrator account by sending a crafted HTTP POST request with the X-Requested-With: XMLHttpRequest header to bypass the CanInstall middleware redirect check. Attackers can supply arbitrary name, email, and password values to the admin-config-setup endpoint, which performs an unauthenticated updateOrInsert targeting the hardcoded administrator user ID, enabling full administrative access to all CRM data.

OpenEMR through 8.2.0 contains an authenticated SQL injection vulnerability in the backup configuration import feature that allows administrators with admin or super ACL privileges to execute arbitrary DDL and DML statements against the application database by uploading a crafted SQL file at the form_step=202 parameter in backup.php. Attackers can exploit the unfiltered shell_exec invocation of the mysql command-line client to extract credential hashes, modify access control tables, inject backdoor accounts, create persistent triggers or stored procedures, and write arbitrary files to the filesystem where MySQL FILE privileges and permissive secure_file_priv settings are configured.

Ghidra contains an arbitrary code execution vulnerability in the Swift demangler analyzer that allows an attacker to execute arbitrary binaries by supplying a malicious Ghidra project with a crafted Swift tool directory path. When a victim opens the attacker-supplied project, SwiftDemanglerAnalyzer restores the persisted Swift binary directory from project state and SwiftNativeDemangler executes the resolved binary without integrity or signature verification, causing attacker-controlled executables to run under the Ghidra process user with no prompt or confirmation.

A flaw was found in GNU tar. When extracting an archive with the --one-top-level option, hardlink targets are not confined to the designated top-level directory and may resolve relative to the extraction working directory. A crafted archive can create hardlinks that escape the intended boundary and, when combined with a preexisting symbolic link under the working directory, may allow writing outside that boundary during a single extraction.

XML::Sig versions before 0.71 for Perl allow XPath injection in ID lookup. verify() and _get_signed_xml() in lib/XML/Sig.pm build XPath expressions by concatenating the SignedInfo/Reference/@URI value read from the document being verified. The value is neither escaped nor checked against the NCName grammar that XML requires of an ID, so a URI containing a single quote closes the string literal in the generated expression and appends arbitrary XPath operators. A crafted URI can make the lookup match elements the reference does not name, or every element in the document, so which node is selected for digest verification is decided by the injected expression rather than by the reference.

xml\

OpenWrt luci-app-dockerman (LuCI master and openwrt-25.12 snapshots containing the ucode docker_rpc.uc RPC backend after the JS/ucode conversion) contains an OS command injection vulnerability. The package's read ACL grants broad ubus access to docker.* / docker.container.*, which exposes the docker.container.ttyd_start method even though it performs mutating operations. The run_ttyd handler builds a shell command from the request-controlled id, cmd, and uid fields and passes it to system() without quoting or argv-style execution in the rpcd root context. An authenticated attacker holding only the luci-app-dockerman read ACL can inject shell metacharacters (e.g., in id) to execute arbitrary commands as root via an HTTP POST to /ubus. openwrt-24.10 and openwrt-23.05 do not contain this backend and are not affected; no patched version was known as of the advisory.

Admidio before 5.0.11 does not validate the adm_csrf_token in modules/category-report/preferences.php, which performs persistent Category Report configuration changes based on GET parameters (delete and copy). An attacker can trick an authenticated administrator into visiting a crafted URL to delete or duplicate Category Report configurations, affecting the integrity and availability of that module's configuration.

Grav CMS versions 2.0.7 through 2.0.10 fail to validate fully-qualified static method calls (Class::method) in blueprint dynamic-field directives because Blueprint::isSafeDynamicCall() only applies its dangerous-callable denylist to strings that do not contain '::'. An account with only page-editing rights (admin.pages, not super-admin or admin.pages_twig) can plant a directive in a page's form-field frontmatter that invokes an arbitrary public static PHP method with attacker-controlled arguments. Using built-in gadget methods this allows reading of any server-readable file (disclosed to anonymous visitors of the crafted page) and arbitrary creation/copying of files and directories under the web-server account. Fixed in 2.0.11.

CTI-Transmute is affected by a server-side request forgery vulnerability in the evaluation report PDF-generation functionality. User-controlled CTI content, including conversion names, descriptions, and comments, is converted from Markdown to HTML and rendered as a PDF using WeasyPrint. Before the patch, the renderer used WeasyPrint’s default URL-fetching behavior without restricting the protocols or destinations that could be referenced by the generated HTML. An attacker able to supply content included in an evaluation report could inject crafted resource references using schemes such as http://, https://, or file://. When the report was rendered, CTI-Transmute could fetch these resources using the application server’s network connectivity and filesystem privileges. Successful exploitation could allow an attacker to: * access services available only from the CTI-Transmute server or its internal network; * probe internal hosts and service endpoints; * retrieve local files readable by the application process; and * expose fetched content through the generated PDF, depending on the referenced resource type and rendering context. The vulnerability is corrected by providing WeasyPrint with a restrictive URL fetcher that permits only self-contained data: URIs. The externally hosted Google Fonts stylesheet was also removed so that PDF generation performs no intentional network or filesystem fetches.

A vulnerability in Wapt Server before version 2.6.1.17813 allows a  remote unauthenticated attacker to bypass security restriction using a specially crafted packet and retrieve a valid session token for the targeted account.

A flaw was found in SSSD. The sss_nss_protocol_parse_addr() function in the NSS responder does not validate the addrlen field against the remaining packet body size. A local attacker can exploit this via a crafted GETHOSTBYADDR request to the NSS responder socket, causing an out-of-bounds read and process crash, resulting in a denial of service.

eParakstītājs 3.0 for Windows before version 1.10.0 retrieves and executes its automatic updates over a channel that is not authenticated or integrity-protected. On each launch the application fetches an update descriptor (XML) over TLS but accepts any TLS certificate (a permissive TrustManager and a HostnameVerifier that always returns true), does not verify any digital signature on the update descriptor, and does not verify the Authenticode signature or a checksum of the downloaded installer before running it. A man-in-the-middle attacker able to redirect www.eparaksts.lv can serve a crafted update descriptor pointing to an attacker-controlled executable, which the client downloads and executes, resulting in arbitrary code execution on the victim host.

A vulnerability in the iDTM FDI allows an attacker with elevated privileges and access to the host system to enable the debug interface by placing a crafted file in the application directory, potentially resulting in unauthorized access to connected devices and exposure, modification, or disruption of device data or operation.

A flaw was found in GIMP. A remote attacker could exploit this by tricking a user into opening a specially crafted PAA (Paint Shop Pro Array) image file. This vulnerability, a heap-based out-of-bounds write in the decode_lzss() function of the PAA file format plugin, allows data to be written beyond the intended memory buffer. This could lead to heap metadata corruption and potentially enable the attacker to execute arbitrary code on the affected system.

In display, there is a possible memory corruption due to use after free. This could lead to local escalation of privilege if a malicious actor has already obtained the System privilege. User interaction is not needed for exploitation. Patch ID: ALPS11019722; Issue ID: MSV-7759.

A vulnerability in huggingface/transformers versions <=5.8.0.dev0 allows an attacker to perform arbitrary file writes via path traversal. The issue resides in the `save_pretrained()` methods of `PreTrainedTokenizerBase` and `ProcessorMixin`, where keys from the `chat_template` dictionary are used directly as filenames without proper validation. An attacker can exploit this by publishing a malicious Hugging Face Hub repository with a crafted `tokenizer_config.json` file. When a victim downloads and saves the tokenizer or processor, the attacker-controlled keys can escape the intended save directory, enabling arbitrary file writes with attacker-controlled content. This vulnerability affects multiple processors inheriting from `ProcessorMixin`, including Idefics, Florence, Gemma, Phi, and Qwen-VL.

Zephyr's Bluetooth Mesh subnet key management leaks one PSA Crypto key slot on every subnet-key teardown. In subsys/bluetooth/mesh/subnet.c, net_keys_create() imports the Private Beacon Key into a PSA key slot under CONFIG_BT_MESH_PRIV_BEACONS (enabled by default), but subnet_keys_destroy() guarded the matching psa_destroy_key() with CONFIG_BT_MESH_V1d1. That Kconfig symbol was removed when explicit Mesh 1.0.1 support was dropped, so the destroy branch became permanently dead code and the import is never balanced by a destroy. The imbalanced teardown is reached every time subnet keys are destroyed: deleting a subnet (Config Server NetKey Delete), completing a Key Refresh Procedure (which retires the old key set), and resetting/re-provisioning the node. The over-the-air triggers are processed only under the node's device key, so they are exercisable by the provisioner or network administrator that owns the node, reachable over the Bluetooth Mesh network. With the default CONFIG_MBEDTLS_PSA_KEY_SLOT_COUNT of 16, repeated add/delete or key-refresh cycles exhaust the shared PSA key-slot pool after roughly a dozen rounds. Once exhausted, bt_mesh_private_beacon_key() and thus subnet creation fail: the node can no longer add subnets or complete key refresh, and other PSA crypto consumers on the device may be starved, until the device is rebooted. The fix aligns the destroy guard with the import guard (CONFIG_BT_MESH_PRIV_BEACONS) so each slot is freed.

better-auth (npm) before 1.4.2 allows an external request to configure baseURL when it is not otherwise defined (e.g., BETTER_AUTH_URL is unset). An attacker able to make the very first request to the server after startup can poison the router's base path, causing all routes to return 404 for all users (denial of service). The issue is not reachable when baseURL is explicitly configured or on typical managed hosting platforms.

better-auth passkey versions before 1.4.0 contain an insecure direct object reference vulnerability in the passkey deletion endpoint that allows authenticated users to delete arbitrary passkeys by ID. Attackers with valid sessions can submit crafted requests to the delete-passkey endpoint with enumerated passkey IDs to remove other users' passkeys.