CVE Catalog

A vulnerability allowing incorrect authorization has been identified in ABB T-MAC Plus versions 4.0 through 24. This issue could potentially be exploited to bypass authorization mechanisms, leading to unauthorized actions or access within the application.

A cross-site scripting vulnerability has been identified in ABB T-MAC Plus versions 4.0 through 24. This issue arises from improper handling of input during web page generation, allowing for the injection of malicious scripts.

A vulnerability in ABB T-MAC Plus versions 4.0 through 24 allows for authorization bypass through user-controlled keys. This could potentially be exploited to manipulate authorization processes, leading to unauthorized actions or access within the application.

A vulnerability exists in ABB T-MAC Plus versions 4.0 through 24, allowing files or directories to be accessed by external parties.

A vulnerability exists in MLflow versions prior to 3.11.0, where AI Gateway secrets can resolve environment variable references. This flaw allows low-privileged authenticated users in basic-auth deployments, or unauthenticated users in default deployments without basic-auth, to exfiltrate sensitive server-side environment credentials to an attacker-controlled endpoint. The issue stems from the 'api_key' field in gateway secrets accepting '$ENV_VAR' references, which are resolved at runtime and sent in authentication headers to the specified 'api_base'. Exploitation could lead to the leakage of critical credentials, such as cloud artifact credentials, potentially causing artifact poisoning and cross-boundary code execution in downstream environments.

A reflected cross-site scripting vulnerability has been identified in the Fox-Themes Prague WordPress plugin, affecting versions through 2.2.8. This vulnerability allows attackers to inject malicious scripts that are executed when users visit the affected site.

A log forging vulnerability has been identified in the Morgan logging middleware, specifically in versions 1.2.0 through 1.10.1. The issue arises because the ':remote-user' token extracts the Basic authentication username from the Authorization header and logs it without neutralizing control characters. This flaw allows an unauthenticated attacker to send a crafted Authorization header containing CR or LF bytes, injecting forged log lines that disrupt the standard one-request-per-line format of access logs. The vulnerability affects the built-in 'combined', 'common', 'default', and 'short' log formats, as well as any custom format that includes ':remote-user'.

A vulnerability exists in Vinyl Cache versions prior to 9.0.1 and Varnish Cache versions prior to 9.0.3, as well as in Varnish Cache releases from 7.6.0 up to and including 8.0.1, and in the Varnish Cache 6.0 LTS series from 6.0.14 up to and including 6.0.17. The issue arises from a deficiency in HTTP/2 request parsing, which can be exploited to launch a backend request desynchronization attack, commonly known as request smuggling. This exploitation can lead to cache poisoning, authentication bypass, and potentially allow for information disclosure and manipulation. The vulnerability is only exploitable if HTTP/2 support is enabled, which is not the default setting.

A buffer overflow vulnerability has been identified in FreeIPMI versions prior to 1.16.18, specifically within the ipmi-oem component. This vulnerability arises from improper handling of response messages in certain subcommands, leading to exploitable buffer overflows. The affected subcommands are 'ipmi-oem dell get-active-directory-config' and 'ipmi-oem fujitsu get-sel-entry-long-text'. When these commands are executed, a malicious or compromised BMC (Baseboard Management Controller) can send responses that exceed the allocated buffer size, potentially corrupting stack memory and causing the ipmi-oem process to crash. Depending on the system's compiler options and runtime mitigations, this stack corruption could be exploited to hijack control flow and execute arbitrary code.

A resource consumption vulnerability has been identified in Dask versions through 3.0, specifically within the 'nunique_approx' function of the 'dask/dataframe/hyperloglog.py' file. This issue arises in the HyperLogLog (HLL) Handler component, where the handling of hash values leads to increased resource usage. The vulnerability can be exploited remotely, although it requires a high level of complexity and is considered difficult to execute.

A SQL injection vulnerability has been identified in SourceCodester Pizzafy E-Commerce System version 1.0. The issue arises in the administrative control panel, specifically within the login function of the file 'admin_class_novo.php'. The vulnerability allows remote attackers to manipulate the 'username' POST parameter, injecting malicious SQL that could be executed by the database. This exploitation bypasses authentication, granting unauthorized access to administrative privileges.

A use-after-free vulnerability has been identified in EIPStackGroup OpENer versions prior to 2.3.0. The issue arises in the 'CreateMessageRouterRequestStructure' function within 'cipmessagerouter.c', specifically in the 'SendRRData' handler. This vulnerability allows for remote exploitation, where a crafted EtherNet/IP explicit-message packet can trigger a stack-use-after-return error, leading to potential memory corruption.

A denial-of-service vulnerability has been identified in Cpanel::JSON::XS versions prior to 4.41 for Perl. The issue arises when input prefixed with a UTF-8 Byte Order Mark (BOM) is processed by a decode filter callback that throws an exception. The decode_json() function attempts to remove the BOM by adjusting the input scalar's string pointer, but if an exception occurs, this adjustment is not reversed. As a result, the scalar is left with an invalid string pointer and a shortened length, leading to a crash when the scalar is freed. This vulnerability can be reliably exploited by decoding a BOM-prefixed document with a filter callback that raises an error, causing the interpreter to abort.

A type confusion vulnerability has been identified in Cpanel::JSON::XS versions prior to 4.41 for Perl. This vulnerability arises when duplicate object keys are processed with the 'dupkeys_as_arrayref' option enabled. The 'decode_hv()' function mishandles these duplicates, leading to a crash. Specifically, the function collapses duplicate keys into an array reference but fails to properly check if the existing value is a reference before dereferencing it. As a result, a non-reference scalar can be incorrectly treated as a reference, causing a segmentation fault. This vulnerability can be exploited by decoding untrusted JSON data with 'dupkeys_as_arrayref' enabled, particularly when the JSON contains duplicate keys.

A local file inclusion vulnerability has been identified in SourceCodester Online Food Ordering System version 2.0. The issue arises in the '/index.php' and '/admin/index.php' files, where user input from the 'page' parameter is directly passed to the PHP 'include' function without proper sanitization or path restrictions. This vulnerability allows remote attackers to include arbitrary PHP files or use PHP stream wrappers to read the source code of sensitive files, such as database configuration files. No login or authorization is required to exploit this vulnerability.

A broken access control vulnerability has been identified in SourceCodester Online Boat Reservation System version 1.0. This vulnerability allows low-privileged users to access administrative functionalities without proper authorization. The issue arises from inadequate authorization checks on administrative endpoints, enabling unauthorized users to modify boat information that should be restricted to administrators.

A Cross-Site Request Forgery (CSRF) vulnerability has been identified in the EmergencyWP WordPress plugin, specifically in versions through 1.4.2. The issue arises from inadequate nonce validation in the settings save handler, allowing unauthenticated attackers to manipulate various plugin settings. This includes changing the minimum access role, data-erasure-on-uninstall flag, life-check timing values, mandatory email address, confirmation page ID, and date/time formats. Exploitation requires tricking a site administrator into clicking a link that initiates the forged request.

A stored cross-site scripting vulnerability has been identified in the Passeum Ticketing plugin for WordPress, affecting all versions through 1.0. The issue arises because the 'get_shop_url()' method returns the 'shop_name' setting value without proper sanitization when it starts with 'http'. This, coupled with inadequate validation in the 'validate_shop_name()' function, which only checks for empty values and string types, allows authenticated attackers with Administrator-level access to inject arbitrary external scripts. By setting the 'shop_name' to a URL controlled by the attacker, the plugin can be made to enqueue JavaScript and CSS from that domain. The injected scripts then execute on the frontend for all visitors, but this vulnerability does not impact single-site installations as administrators have the 'unfiltered_html' capability.

A regular expression denial-of-service (ReDoS) vulnerability has been identified in johnhuang316 code-index-mcp versions through 2.14.0. The issue arises in the 'search_code_advanced' component, where the 'is_safe_regex_pattern' function fails to properly validate user-supplied regular expression patterns. This oversight allows for the introduction of inefficient regex patterns that can be exploited to cause excessive backtracking, leading to significant CPU exhaustion. The vulnerability can be exploited remotely, especially when the application is running without external search tools like ripgrep, ag, or grep, which are necessary to disable the vulnerable regex handling.

A Regular Expression Denial-of-Service (ReDoS) vulnerability has been identified in Wonderwhy-er DesktopCommanderMCP versions through 0.2.38. The issue arises in the 'start_search' component, specifically within the 'src/search-manager.ts' file. The vulnerability allows for inefficient regular expression processing that can be exploited remotely, causing the application to hang and become unresponsive. This issue has been publicly disclosed and can be exploited by injecting a malicious regex pattern through an AI assistant integrated with the Desktop Commander MCP server.

A server-side request forgery (SSRF) vulnerability has been identified in Wonderwhy-er DesktopCommanderMCP version 0.2.37. The issue arises in the 'readFileFromUrl' function within 'src/tools/filesystem.ts', where user-supplied URLs are sent directly to the 'fetch()' function without proper validation. This flaw allows remote attackers to manipulate the URL argument and make the server perform unauthorized requests to internal services or cloud metadata endpoints, potentially leading to data exfiltration.

A vulnerability in LibreChat versions through 0.8.3 allows a shared-agent editor to globally delete file records via the DELETE /api/files endpoint. This deletion affects all agents referencing the same file_id, including private agents of the file owner, which are left with a broken reference to the deleted file. The issue arises because the file deletion is not scoped to the requesting agent, leading to unintended cross-agent impacts.

A vulnerability in LibreChat versions through 0.8.3 allows users with VIEW access to an MCP server to access decrypted admin-managed secrets. This is achieved through the 'GET /api/mcp/servers' and 'GET /api/mcp/servers/:serverName' endpoints, which return plaintext values for 'apiKey.key' and 'oauth.client_secret'. The issue arises because the MCP registry decrypts sensitive server configurations before sending them in the response, without redacting these secrets. As a result, viewers can exfiltrate provider credentials that should remain confidential.

A vulnerability exists in the Go programming language's net/textproto package, where functions include arbitrary input in error messages without proper escaping. This flaw could enable an attacker to inject misleading content, such as terminal control bytes, into errors that are printed or logged. The issue affects Go versions prior to 1.25.11 and from 1.26.0 up to but not including 1.26.4.

A denial-of-service vulnerability has been identified in the Go programming language's standard library, specifically within the MIME package. The issue arises in the WordDecoder.DecodeHeader function, which can be subjected to a denial-of-service attack by decoding a maliciously-crafted MIME header. This header can contain numerous invalid encoded-words, leading to excessive CPU consumption. The vulnerability affects Go versions prior to 1.25.11, as well as versions from 1.26.0 up to but not including 1.26.4.

A vulnerability exists in alf.io versions through 2.0-M5-2509-1, allowing authenticated administrators to read arbitrary files accessible to the JVM process user. The issue arises from the extension sandbox injecting a fully-functional HTTP client, 'simpleHttpClient', into every extension script. The 'postFileAndSaveResponse()' method, which lacks path validation, directory restrictions, and allowlists, can be exploited to exfiltrate file contents to an attacker-controlled server via HTTP POST. This vulnerability could lead to the exposure of sensitive information such as database credentials, TLS private keys, SSH private keys, cloud credentials, and application secrets.

A stored cross-site scripting vulnerability has been identified in GLPI versions 11.0.0 prior to 11.0.6. This issue allows a technician to inject an XSS payload into ITIL costs, which could be executed later when the data is viewed. The vulnerability has been addressed in GLPI version 11.0.7.

A sandbox escape vulnerability has been identified in alf.io, an open-source ticket reservation system, prior to version 2.0-M5-2606. This vulnerability allows authenticated administrators to execute arbitrary operating system commands on the server. The issue arises in the extension script engine, which is designed to run restricted JavaScript in a sandboxed Rhino environment. However, an unguarded injected Java object, 'returnClass', combined with an incomplete Abstract Syntax Tree (AST) blocklist, enables a full escape from the sandbox using Java reflection, bypassing validation errors. The vulnerability is triggered when an extension script is saved, with the malicious code executing during the next relevant event, such as 'INVOICE_GENERATION' on the first ticket purchase with invoicing enabled.

A critical information disclosure vulnerability exists in LibreChat versions through 0.8.3, specifically within the Model Context Protocol (MCP) server integration. The issue arises because the validation schema for MCP server URLs resolves `${VAR}` placeholders using the server's environment variables during user input validation. This flaw allows any authenticated user to create a malicious MCP server configuration that, when activated, causes the LibreChat server to connect to an attacker-controlled domain and leak sensitive information such as the CREDS_KEY, CREDS_IV, JWT_SECRET, and MONGO_URI via the request URL. The vulnerability enables a full compromise of the application's cryptographic materials and database credentials, all without requiring administrative rights.

A vulnerability allowing Insecure Direct Object Reference (IDOR) has been identified in LibreChat versions through 0.7.6. This vulnerability exists in the API keys management endpoint (PUT /api/keys), where authenticated users can inject a userId parameter to overwrite other users' API keys for providers like OpenAI, Anthropic, and Azure. The issue arises because the JavaScript object spread operator is applied after the authenticated user's ID is set, allowing for unauthorized modification of API key configurations. This vulnerability has been patched in version 0.8.3-rc1.

A performance issue has been identified in the Go programming language's x509 certificate handling, specifically in the VerifyHostname method. This vulnerability arises because VerifyHostname processes each DNS Subject Alternative Name (SAN) entry in a loop, repeatedly splitting the hostname into labels. As a result, the verification time can increase significantly, especially with a large number of SAN entries, leading to excessive CPU usage. This problem is exacerbated when an application passes an attacker-controlled hostname without proper validation, allowing for even longer inputs that amplify the issue. Notably, this overhead occurs even with untrusted certificates, as hostname verification is performed before establishing the certificate chain.

A vulnerability exists in QloApps versions through 1.7.0, related to weak password hashing. The Tools::encrypt() function in classes/Tools.php uses MD5 to hash passwords, combining them with a static cookie key. This allows attackers to conduct offline brute-force attacks on the MD5 hashes. The issue is exacerbated by the automatic generation of 8-character passwords for guest-to-customer account conversions, as detailed in classes/Customer.php, making it easy to recover credentials.

A vulnerability allowing an out-of-bounds write has been identified in Seagate's openSeaChest version 25.05.3, across all supported platforms. This issue arises in the '--showSupportedFormats' function, where the software improperly manages memory when interacting with NVMe devices. Specifically, it allows for writing an extra byte outside of allocated memory by exploiting a bogus value in the namespace FLBAS byte. This mismanagement could potentially be leveraged to manipulate memory in a harmful way.

A vulnerability allowing out-of-bounds write has been identified in Seagate's openSeaChest version 26.03.0, across all supported platforms. This issue arises during the Trim/Unmap operation, where the software improperly manages memory allocation. Specifically, it allows for writing additional data that describes a range of logical block addresses (LBAs) to deallocate 16 bytes beyond the allocated memory, potentially leading to memory corruption.

A vulnerability in Seagate's OpenSeaChest version 25.05.3, across all supported platforms, allows for out-of-bounds write and read operations. This issue arises in the '--showSCSIDefects' function, where defect information can be improperly written for large defect lists. The vulnerability can be triggered by a severely defective drive or a maliciously crafted SCSI device that manipulates the defect response length.

A code injection vulnerability has been identified in AhujaSid's Blender-MCP integration, specifically in versions up to commit 7636d13bded82eca58eb93c3f4cd8708dfdfbe8b. The issue arises in the 'execute_blender_code' function within 'src/blender_mcp/server.py', where user-controlled input is executed via Python's 'exec()' function without any sanitization. This vulnerability allows for remote code execution on the system running Blender, as the 'exec()' function can be used to execute arbitrary Python code with full access to the Python standard library and, potentially, the underlying operating system.

A server-side request forgery (SSRF) vulnerability exists in the AhujaSid Blender-MCP integration, specifically within the MCP server component. The issue arises because the 'zip_file_url' parameter in the 'import_generated_asset_hunyuan' tool is passed to the 'requests.get()' function without proper validation, allowing the Blender process to make HTTP GET requests to arbitrary destinations. This vulnerability can be exploited remotely, potentially leading to unauthorized access to internal services or metadata endpoints. The flaw has been publicly disclosed and could be actively exploited.

A vulnerability in the gRPC FUSE kernel module of Docker Desktop can cause a virtual machine panic due to unbounded recursion. This issue occurs when a container creates deeply nested directories in a bind-mounted host folder, triggering a dentry invalidation event. The problem has been addressed in Docker Desktop version 4.76.0.

A cross-site scripting (XSS) vulnerability has been identified in OpenCTI versions prior to 7.260227.0. The issue arises in the rendering of email-message observable body data, where the content is not properly sanitized. This vulnerability requires user interaction but could be exploited by individuals sharing STIX or using any of the ingesters. The consequence of this vulnerability could be cross-site request forgery (CSRF) leading to large-scale session theft.

An arbitrary file read vulnerability has been identified in the AhujaSid Blender MCP integration, specifically within the Hunyuan3D model generation feature. This vulnerability arises from the MCP server's handling of the 'input_image_url' parameter, which is passed to the Blender addon without proper validation. Local file paths can be exploited to read sensitive files, such as '/etc/passwd', and exfiltrate their contents to an external API endpoint. The issue has been publicly disclosed and is present in versions of the Blender MCP addon prior to the latest commit, which introduced necessary validation measures.

A denial-of-service vulnerability has been identified in Warmcat Libwebsockets versions prior to 4.5.8. The issue resides in the SSH protocol handler, specifically within the function 'lws_ssh_parse_plaintext' in 'plugins/protocol_lws_ssh_base/sshd.c'. The vulnerability arises from a lack of proper validation on the 'msg_len' parameter, allowing for unbounded memory allocation. This flaw can be exploited remotely, leading to excessive resource consumption and potentially causing the server process to be terminated by the operating system's out-of-memory killer.

A local security vulnerability has been identified in Dräger Zeus Infinity Empowered (Zeus IE) and Zeus RS C500 anesthesia workstations. This vulnerability allows unauthorized individuals with physical access to manipulate the software integrity of the devices through the unprotected USB interfaces. Exploitation of this vulnerability could disrupt therapy functions, alter device-processed data, or use the workstation as a gateway for wider network-based attacks, especially when connected to a network or Dräger Service Connect.

A denial-of-service vulnerability has been identified in Dräger Core version 1.0.5 and Dräger M540 Converter Service version 1.0.9. This vulnerability allows network-adjacent attackers to cause high CPU load by sending specially crafted, unencrypted SDC messages during the discovery process. Attackers with access to the hospital network can exploit this issue by sending malformed SDC packets that exhaust CPU resources in the affected process, leading to a disruption in the processing of further SDC messages.

A vulnerability has been identified in the Dräger Infinity Acute Care System and Standalone Infinity M540 patient monitors, specifically in versions VG4.1.1, VG4.0.3, and earlier, with VG4.2 being partially affected. This vulnerability involves network message handling, allowing remote attackers to inject spoofed or tampered data, leading to denial-of-service conditions. Exploitation of this vulnerability can disrupt network communications, enabling attackers to modify device settings such as alarm states or limits. Additionally, the vulnerability can be exploited to overwhelm the system with excessive network traffic, causing the Cockpit or M540 monitor to reboot and lose network functionality.

A local privilege escalation vulnerability has been identified in Dräger Protector Software versions prior to 6.4.2. This vulnerability arises from insecure file system permissions, allowing local attackers to execute arbitrary code with elevated privileges. Exploitation involves replacing binaries or loaded modules on the host system to run code with NT SYSTEM privileges.

A local privilege escalation vulnerability has been identified in Dräger Protector Software versions prior to 6.4.2. This vulnerability arises from insecure file system permissions, allowing local attackers to execute arbitrary code with elevated privileges. Exploitation involves replacing binaries or loaded modules on the host system to run code with NT SYSTEM privileges.

A vulnerability in authentik, an open-source identity provider, allows for bypassing the Source stage in authentication flows. This issue is present in versions through 2025.12.5, 2026.2.3, and 2026.5.0. The vulnerability arises because the Source stage can be skipped by sending an empty POST request. When the Source stage is active, the flow executor expects a restore token from the user’s interaction with the identity provider (IdP). However, an empty POST request is accepted as valid, allowing the flow to advance without proper authentication. This bypass can be exploited if the Source stage is part of an authentication flow that requires interaction with an IdP.

A vulnerability in authentik, an open-source identity provider, allows an attacker with the ability to modify source connections and an account in one of the configured sources to log into any account. This issue affects authentik versions prior to 2025.12.6, 2026.2.4, and 2026.5.1. The vulnerability arises because the 'UserSourceConnection.user' and 'GroupSourceConnection.group' fields can be changed through the API, enabling low-privilege attackers to misrepresent users or groups and authenticate as them.

A path traversal vulnerability has been identified in BrowserStack Runner versions through 0.9.5. The issue resides in the default HTTP handler within lib/server.js, where the server fails to properly validate file paths before serving them. This vulnerability allows unauthenticated network-adjacent attackers to read arbitrary files from the host filesystem. The HTTP server, bound to all interfaces, can be exploited to traverse outside the project root and access sensitive files.

A remote code execution vulnerability has been identified in BrowserStack Runner versions through 0.9.5. The issue resides in the '/_log' HTTP handler, which accepts POST requests with JSON payloads. The handler processes user-supplied data by passing it to 'vm.runInNewContext()' combined with 'eval()', allowing unauthenticated network-adjacent attackers to execute arbitrary code on the host system. Exploitation involves crafting a JSON request that leverages a host-context Function reference to escape the Node.js vm sandbox and access the host process, achieving full remote code execution without authentication.