CVE Vulnerability Catalog

Vulnerability in StarTraining up to 3.8.1 in the anonymous function of /druid/index.html of the Druid Console component. Missing authentication allows remote attack. Exploit is publicly available.

Vulnerability in StarTraining up to 3.8.1 in the UserLoginService.createToken function of application.yml in the JWT Token Handler component. Manipulation of user_id/company_id causes use of hard-coded password. Remote attack, public exploit.

Vulnerability in Zhonglun CloudPos up to 3.0.1.76 in the OpenLocalBrowser function of ZlPos/ZlPos/Bizlogic/JSBridge.cs in the JSBridge component. Manipulation of the url argument leads to code injection. Remote attack, public exploit.

A vulnerability in golang.org/x/playground allows a malicious txtar file to escape the intended execution context and force arbitrary writes to the playground host's trusted filesystem. Additionally, one of the paths to invoke 'go vet' did not correctly restrict the execution environment, allowing a read of an environment configuration file from the host's $HOME. Together, these can lead to remote code execution on the playground host.

Vulnerability in NetworkManager-iodine, the iodine VPN plugin for NetworkManager. An unprivileged local user can exploit the processing of the 'nameserver' setting when establishing an iodine VPN connection. Embedding shell metacharacters in the 'nameserver' value allows arbitrary command injection and execution with root privileges before dropping elevated permissions.

Vulnerability in BerriAI LiteLLM before 1.101.0-rc.1, allowing tenant isolation bypass in the semantic cache layer. Authenticated users can read other tenants' cached responses by exploiting a metadata key mismatch between _get_semantic_cache_tenant_scope() and _get_metadata_variable_name(). Attackers with a valid virtual key can submit semantically similar prompts to retrieve cached responses containing other tenants' PII, financial data, or source code, and can cause agentic front-ends to auto-execute attacker-supplied tool calls under victim credentials.

Federation upstream in RabbitMQ skips vhost authorization, allowing cross-vhost message access. A policymaker on one vhost can read and drain messages from another vhost it has no permission on, and with the default ack-mode the source messages are consumed rather than copied.

Native MQTT and MQTT over WebSocket behind a trusted PROXY Protocol frontend in RabbitMQ can lose the proxy-derived client address before the MQTT authentication path checks loopback_users, causing the frontend-to-broker address to be treated as loopback. An attacker with valid credentials for a loopback-restricted account can bypass the source-address restriction.

Vulnerability in RabbitMQ from 4.2.0 until 4.3.3 and 4.2.9, exposing the OAuth2 client secret via an unauthenticated JavaScript endpoint. When OAuth2 authentication is enabled for the RabbitMQ Management UI with a client secret, the oauthclientsecret value is included in JavaScript served by the unauthenticated endpoint /js/oidc-oauth/bootstrap.js. Any user who can reach the management UI port can retrieve the OAuth2 client secret.

Incomplete fix for CVE-2026-44838 in RabbitMQ: escaperegexchar/1 does not escape the - character, allowing an MQTT topic permission bypass. When a topic permission template places {clientid} inside a [...] character class, a low-privileged authenticated MQTT user can broaden topic authorization.

The Web STOMP WebSocket handler in RabbitMQ enforces neither max_frame_size nor login_timeout before authentication, allowing an unauthenticated client to keep a connection alive with a slow stream of small frames and accumulate unbounded pre-authentication state. The rabbitmq_web_stomp plugin must be enabled.

RabbitMQ versions from 3.13.0 until 3.13.15, 4.0.20, 4.1.11, and 4.2.6 have a vulnerability in the stream protocol where the FrameMax value negotiated during the Tune handshake is stored but not compared with an inbound frame's declared length before buffering. A remote client can cause excessive memory pressure and denial of service by declaring an oversized frame.

RabbitMQ versions in the 3.13, 4.0, 4.1, 4.2, and 4.3 lines have a vulnerability where fill/2 substitutes ${username} into user_dn_pattern without RFC 4514 DN escaping, allowing a crafted username to alter the LDAP bind DN. Exploitation requires rabbitmq_auth_backend_ldap with a user_dn_pattern containing ${username}, a suitable directory layout, and a valid password. Fixed versions are conflicting across sources.

Zammad before version 7.0.2 has an issue with permission checks in the knowledge base management area. Under certain conditions, data validation for linked items was not fully enforced, potentially allowing users with limited read permissions to interact with items outside their assigned access scope.

Zammad before version 7.0.2 allows a customer who can view a ticket to not see internal ticket articles through the article listing API, but the same customer can directly request an attachment belonging to an internal article via the attachment download endpoint, bypassing article-level authorization. The attachment download only checks the parent ticket, not the article, so the same customer can download the attachment directly.

The Rattadan Cosmowarp smart contract before version 56c6147 can have a comparison to an unintended value of current_admin. This may lead to incorrect contract logic behavior.

InvoicePlane before version 1.7.2: Invoices::generate_xml() appends a database-derived xml_id to the XMLconfigs helper directory and includes the resulting PHP path without validating the identifier. A low-privileged attacker who can influence the e-invoice configuration can use traversal sequences to include an existing PHP file. Code execution requires a separate file-upload or file-write primitive.

InvoicePlane before version 1.7.2 exposes Invoices::delete() and Invoices::delete_invoice_tax() as state-changing routes without requiring POST and validating a CSRF token. When an authenticated administrator loads attacker-controlled content that requests an affected route, the application can delete an invoice or invoice tax record. The cross-origin action can remove financial data without the administrator's intent.

Piwigo before version 16.4.0: check_upgrade_access_rights() in admin/include/functions_upgrade.php conditionally escapes the submitted username only when the removed get_magic_quotes_gpc function exists, so PHP 8 and later concatenate an unauthenticated username directly into the upgrade authentication SQL query. When database upgrades are pending, a crafted query result can satisfy the status and password checks, set PHPWG_IN_UPGRADE, and authorize upgrade execution without valid administrator credentials.

Piwigo before version 16.4.0: admin/element_set_ranks.php stores administrator-controlled image_order[] values without enforcing the existing sort-field whitelist. The stored album image_order expression is later concatenated into ORDER BY clauses by admin/batch_manager_global.php, admin/batch_manager_unit.php, include/section_init.inc.php, and include/ws_functions/pwg.categories.php. When at least one album contains at least one photo, an authenticated administrator can store a crafted expression and trigger it in a later album or Batch Manager query to disclose, modify, or disrupt database data.