A memory handling flaw exists in dnsmasq's DHCPv6 relay processing that can be exploited remotely to cause the service to crash, resulting in a denial of service. This vulnerability affects systems configured as DHCPv6 relays and is rated as moderate impact due to its limited exposure and lack of data compromise. Standard deployments do not enable DHCPv6 relay by default, and older Red Hat platforms do not contain the vulnerable code. Mitigation involves disabling DHCPv6 relay if not needed or restricting network access to the relay port.
A vulnerability in KubeVirt's containerized-data-importer allows tenants with VirtualMachine create permissions to bypass service-account clone authorization checks when using a DataVolumeTemplate with the cdi.kubevirt.io/storage.checkStaticVolume annotation and a cross-namespace clone source. This flaw enables unauthorized cloning of PersistentVolumeClaims (PVCs) from other namespaces, leading to potential cross-tenant data disclosure.
A heap-based buffer overflow vulnerability exists in Corosync's Totem SRP membership commit token handling due to insufficient upper bound checks on message length. An attacker with network adjacency and cluster shared key access can send a crafted oversized commit token to a node in the GATHER membership state, causing heap memory corruption and denial of service. The vulnerability affects Corosync versions prior to 3.2.0 where encryption enforcement is absent or can be disabled. Red Hat Enterprise Linux ships Corosync as part of its High Availability Add-On, which by default enables cluster message authentication, mitigating practical risk. The CVSS v3.1 score is 5.7 (medium severity).
Kartones flask-calendar versions before commit bfde72e have a stored cross-site scripting (XSS) vulnerability. This occurs because the function task_details_for_markup returns unescaped content that is rendered with the |safe filter. Authenticated shared calendar members can inject malicious scripts into task details, which then execute in the browsers of other members, potentially allowing attackers to read and modify private calendar data.
A vulnerability in WhatsApp's remote-attestation verifier for confidential computing could allow bypassing the hardware root of trust by improperly validating the certificate chain. This affects multiple WhatsApp versions on iOS, Android, and Mac Desktop prior to specified releases. However, no exploitation has been observed as all published VCEK certificates properly chain to the hardware root of trust.
Cloud Mail versions prior to 3.6.0 have a stored cross-site scripting (XSS) vulnerability in the attService.addAtt function. This flaw allows unauthenticated attackers to upload inline attachments with attacker-controlled MIME types such as SVG or HTML. By exploiting this, attackers can compute content-hash URLs and steal session JWT tokens from localStorage, enabling account hijacking including administrator accounts.
LessPass self-hosted sync server backend versions before commit 9469b0c lack rate limiting on JWT login endpoints, allowing unauthenticated remote attackers to perform unlimited authentication attempts. This vulnerability enables brute force or credential stuffing attacks against known account emails to obtain tokens with read and write access to stored profiles.
one-hub versions prior to 0.14.28 contain a server-side request forgery (SSRF) vulnerability in the GetImageFromUrl function within common/image/image.go. This flaw allows attackers with a standard non-admin API token to craft requests that fetch image URLs without destination restrictions, potentially enabling access to internal services or cloud metadata endpoints. The vulnerability is fixed in version 0.14.28.
Pydici versions prior to 2026.11.2 contain a stored cross-site scripting (XSS) vulnerability that allows authenticated users with contact edit permissions to inject malicious scripts via contact names. These scripts are rendered unescaped in the contact detail mermaid diagram, leading to script execution in other users' sessions. The vulnerability is fixed in version 2026.11.2.
A critical vulnerability exists in github.com/eclipse-keypont/crypto11/v2 versions up to 2.0.0-rc8 and related deprecated v1 module paths. The issue involves improper nonce length validation in the cipher.AEAD implementation returned by SecretKey.NewGCM(), allowing empty nonces to be used. On tokens like SoftHSM2 that accept zero-length GCM IVs, this leads to keystream reuse and exposure of the GHASH authentication key, enabling forgery attacks. Additionally, configuration options related to GCM IV handling can cause panics or silent encryption under empty IVs. The v1 line also fails to copy the token-generated IV back to the caller, causing nonce mismatches during encryption.
Mattermost versions 11.11.x up to 11.11.1, 11.7.x up to 11.7.14, and 11.10.x up to 11.10.4 have a vulnerability where the broadcast scope of group WebSocket events is not properly restricted. This allows authenticated users, including guests, to receive information about non-referenceable groups such as display name, description, member count, and member IDs via the received_group event when groups are created, updated, deleted, or restored.
Iceburg CRM versions through 5.0.0 have an authorization vulnerability in the Module::saveRecord() function. Authenticated users with write access to one module can exploit this flaw by prefixing request body keys with another module's ID, allowing them to write to unauthorized modules. This can be used to escalate privileges by setting their role_id to Admin, thereby gaining access to administrative data.
A SQL injection vulnerability exists in WhatsApp for Android versions prior to 2.26.35.74. This flaw can be triggered when a recipient taps on a specially crafted “order details” message from an unofficial client. Exploitation could allow deletion of locally stored WhatsApp files and, if the user is online at the time, leak one bit of information about the user's chat database via a timing side-channel. There is no evidence of exploitation in the wild.
The PyPI package 'scope-token-fetch-vector' version 1.0.0 and 1.0.0.post1 contains malicious code that exfiltrates basic system information such as IP address and username upon installation or import. The package overrides the install command in setup.py to execute this malicious behavior. It serves no legitimate purpose beyond this data exfiltration.
Pozeiden versions prior to 0.3.0 contain multiple cross-site scripting (XSS) vulnerabilities due to insufficient escaping and validation of MermaidJS-rendered SVG content. This allows attacker-controlled input to inject script-bearing SVG markup that executes when the SVG is embedded raw into HTML by first-party consumers. The issue is resolved in version 0.3.0.
The lyxe-baileys npm package contains malicious code that compromises any computer on which it is installed or running. The compromise is severe enough that all secrets and keys stored on the affected machine should be rotated immediately from a different, trusted device. Removing the package alone may not fully eliminate the malicious presence, as the attacker may have established persistent control.
Pozeiden is a MermaidJS renderer written in Zig. Prior to 0.3.0, src/renderers/c4.zig, src/renderers/class.zig, src/renderers/er.zig, and src/renderers/requirement.zig use fixed row-layout arrays while deriving subsequent loop bounds from an untrusted diagram's actual item count, and src/renderers/state.zig indexes a fixed depth array using user-controlled nesting depth. A valid diagram with at least 193 grid items can make the row count exceed the 64-row storage, while a deeply nested state diagram can exceed the col_count depth storage through the public pozeiden.render() path. Safe CLI and ReleaseSafe builds abort the rendering process, enabling reliable denial of service in applications that render untrusted diagrams. ReleaseSmall WASM builds with safety checks disabled instead perform out-of-bounds stack reads or writes, creating a memory-safety violation. This issue is fixed in version 0.3.0.
CVE-2026-129995 is an OS command injection vulnerability in bvisible's mcp-ssh-manager prior to version 3.8.5. The flaw exists in backup-related functions that build remote shell commands from user-controlled inputs without properly neutralizing shell metacharacters. This allows attackers with SSH user privileges to execute arbitrary OS commands on managed remote servers. The vulnerability affects commands such as ssh_backup_create, ssh_backup_restore, ssh_backup_schedule, and ssh_backup_list, with ssh_backup_schedule enabling persistent execution via cron. The ssh_backup_list command bypasses server policy checks, allowing command execution even in restricted modes. The issue was fixed in version 3.8.5.
CVE-2026-77383 is a high-severity OS command injection vulnerability in bvisible mcp-ssh-manager versions prior to 3.6.7. Several functions in the database manager module improperly interpolate user-controlled inputs into shell commands without sanitization, allowing arbitrary command execution on SSH targets. The ssh_db_list function remains exploitable even in readonly and restricted modes, bypassing intended security boundaries. This issue was fixed in version 3.6.7.
CVE-2026-106075 is an improper access control vulnerability in Rocket.Chat affecting versions prior to 9.0.0. It allows authenticated users with standard privileges to inject arbitrary message metadata via the chat.sendMessage API endpoint. This can result in persistent denial of service at the room level, stored HTML/iframe injection, unauthorized modification of private starred messages, permission bypass, spoofing of pinned message attribution, forged edit attribution, and invisible user mentions.