Threat briefs

An Elastic machine learning job detects unusually large file transfers by remote hosts, indicating potential lateral movement or data exfiltration by adversaries who consolidate data into single large files to avoid detection.

An Elastic machine learning rule detects potential Domain Generation Algorithm (DGA) activity, commonly used by malware for command and control (C2) communication, by analyzing DNS requests from source IP addresses to identify aggregate patterns indicative of DGA usage.

Elastic has released a machine learning detection rule designed to identify potential data exfiltration attempts by flagging anomalous spikes in the volume of data written to external devices, indicating illicit data copying or transfer activities by threat actors.

A machine learning job by Elastic detects potential data exfiltration by identifying anomalous network traffic patterns where high bytes are sent to an unusual destination port, suggesting data is being exfiltrated via command and control channels.

Elastic's machine learning rule detects potential data exfiltration by identifying anomalous network traffic, specifically large data transfers to unusual geo-locations via IP addresses, indicating possible exfiltration over command and control channels.

Elastic Security's statistical model identifies command-and-control (C2) beaconing activity in network logs on Windows and Linux systems by analyzing network traffic patterns and excluding known benign processes, enabling defenders to detect and respond to stealthy adversary communications for persistence and data exfiltration.

A path traversal vulnerability, tracked as CVE-2026-67185, exists in TinyWeb through version 0.0.8, allowing unauthenticated attackers to read arbitrary files by submitting '..' sequences in the URL path, bypassing security checks and potentially exposing sensitive data like credential stores or private keys when the server runs with root privileges.

A null pointer dereference vulnerability, CVE-2026-67184, in TinyWeb through version 0.0.8 allows unauthenticated remote attackers to crash worker processes by sending a malformed HTTP request line with an invalid version string, leading to a denial of service.

A critical memory leak vulnerability, CVE-2026-67183, in TinyWeb versions up to 0.0.8 allows unauthenticated attackers to exhaust server memory by sending ordinary HTTP requests, leading to worker process termination and denial of service.

An HTTP request smuggling vulnerability, identified as CVE-2026-67182, in Rouille versions 0.3.3 through 3.6.2 allows remote attackers to bypass access controls by injecting bare line feed characters (0x0A) into client-supplied request header values, causing upstream backends to misinterpret subsequent data as a separate, attacker-controlled HTTP request.

A vulnerability, CVE-2026-16313, exists in the `sg_inq` command of `sg3_utils` on Red Hat Enterprise Linux systems, allowing an attacker who can present a specially crafted SCSI device to inject arbitrary properties into the `udev` device database by embedding a newline character in the device's name string, leading to arbitrary command execution as root when the device is disconnected.

The pytonapi library, specifically version 2.2.0, contains an authentication bypass vulnerability (GHSA-3fcr-jvgp-7f58) in its TonapiWebhookDispatcher, allowing unauthenticated remote attackers to send forged payloads to custom webhook endpoints, triggering victim-defined business logic and causing integrity impact.

A denial-of-service vulnerability (CVE-2026-54632) exists in the SIPSorcery NuGet package versions <= 10.0.8, allowing an unauthenticated attacker to remotely terminate an active RTP or WebRTC media session by sending a single malformed inbound UDP packet to the RTP/ICE socket, which exploits insufficient length checks and an exception handling flaw.

Poweradmin versions earlier than 4.2.4 and from 4.3.0 up to, but not including, 4.3.3 are vulnerable to CVE-2026-54588, a critical Host Header Injection flaw in OIDC, SAML, and logout authentication flows that allows an unauthenticated attacker to manipulate the HTTP_HOST header, poisoning callback URLs to redirect authorization codes to an attacker-controlled server, leading to full account takeover and potential full DNS zone control.

An unauthenticated attacker can exploit an unbounded RECONNECT_REQUEST forwarding vulnerability in QTINeon's NeonRelay component to amplify denial-of-service attacks against a connected host. By sending spoofed RECONNECT_REQUEST packets, the relay forwards each one to the host without proper deduplication or rate limiting, consuming host resources. Additionally, excessive spoofed IPs can reset legitimate rate limiting, further impacting service availability. This vulnerability affects Java, Python, and TypeScript implementations of NeonRelay.

The `OAuth2::Client` in the `oauth2` Ruby gem is vulnerable to credential disclosure and Server-Side Request Forgery (SSRF) due to improper handling of protocol-relative redirect URLs, allowing an attacker to steal bearer tokens and access internal network resources.

The 'oauth' Ruby gem versions 0.5.5 through 1.1.5 are vulnerable to a critical issue (CVE-2026-54605) where the 'OAuth::Consumer#token_request' method improperly handles HTTP 3xx redirects during OAuth 1.0 token exchanges, enabling an attacker to redirect the request to a malicious host, exposing sensitive OAuth 1.0 metadata, and facilitating Server-Side Request Forgery (SSRF) and confused-deputy behavior.

Rouille HTTP server framework versions 0.3.3 through 3.6.2 are vulnerable to an HTTP request smuggling attack, CVE-2026-67181, allowing remote attackers to desynchronize HTTP message boundaries by exploiting improper header forwarding in the proxy implementation, leading to potential bypassing of security controls or unauthorized access.

A reachable assertion vulnerability exists in the `Request::remove_prefix` function of the Rouille web server framework, affecting versions 0.1.6 through 3.6.2, allowing remote, unauthenticated attackers to crash the server and cause a denial of service by sending a crafted percent-encoded URL.

Camaleon CMS versions 2.1.1 through 2.9.1 are vulnerable to authenticated remote code execution where an attacker with `custom_fields manage` permission can execute arbitrary Ruby code by injecting a malicious expression into the `select_eval` custom field type's options command parameter, which is then evaluated via `instance_eval` within an ERB view when a post edit page is rendered, leading to server-side code execution with web server process privileges.

Unauthenticated Model Context Protocol (MCP) servers, particularly those running protocol version 2024-11-05, are widely exposed across cloud environments, enabling significant security risks by allowing attackers to bypass authentication, gain initial access, execute arbitrary commands on backend systems, obtain sensitive cloud credentials (including temporary ones via Server-Side Request Forgery against cloud metadata endpoints), discover internal systems and data, and collect/exfiltrate sensitive information like PII, business records, and security findings.

A privilege escalation vulnerability, CVE-2026-54593, exists in Pterodactyl's Wings component that allows authenticated subusers to upload arbitrary files to a server without explicit file creation permissions, due to insufficient validation of panel-signed JSON Web Tokens (JWTs.

An Uncontrolled Search Path Element vulnerability, CVE-2026-8164, in ArkSigner Desktop Client versions from v2.2.16.10 through 17062026 allows a local attacker to perform Search Order Hijacking, potentially leading to arbitrary code execution or privilege escalation with the application's privileges.

An improper privilege escalation vulnerability (CVE-2026-63727) exists in Anchore Enterprise versions 5.11.0 to 5.27.1 and 6.0.0, specifically within the user management API, allowing an authenticated attacker to issue a crafted API call to modify user permissions and gain elevated access to resources and operations, such as granting write access to a read-only user, with fixes available in versions 5.27.2 and 6.0.1.

An unauthenticated attacker can exploit CVE-2026-61609, a global rate-limit vulnerability in Pterodactyl Panel versions up to and including 1.12.4, by sending approximately 10 requests per minute to authentication endpoints, leading to a panel-wide denial of service for all legitimate users and administrators attempting to log in or complete 2FA.

@wakaru/cli versions from 1.0.0 up to, but not including, 1.4.0 are vulnerable to arbitrary file write due to a path traversal flaw when unpacking a crafted JavaScript bundle using the `--unpack` command, where specially formatted filenames can bypass sanitization and lead to remote code execution.

A nil pointer dereference vulnerability, tracked as CVE-2026-47427, in the GitHub MCP Server's `completion/complete` handler allows an unauthenticated attacker to cause a complete denial of service by sending a malformed JSON-RPC request with missing or empty parameters for the `ref` field, leading to an immediate server crash.

WordPress Coding Standards (WordPressCS) versions before 3.4.1 are vulnerable to arbitrary code execution due to a flaw in the `WordPress.WP.EnqueuedResourceParameters` sniff, allowing an attacker to execute arbitrary commands on the scanning host by crafting a malicious `$ver` argument, posing a risk for users running PHPCS with specific rulesets in CI pipelines or developer environments.

An inverted-boolean bug (CVE-2026-46428) in the `lettre` library's `boring-tls` integration silently disables TLS hostname verification for callers using the default strict configuration, allowing an on-path attacker with any chain-valid certificate to intercept SMTP submission, including credentials and message contents, from affected `lettre` clients.

A vulnerability, CVE-2026-43910, in Appium's java-client allows a malicious Appium server to redirect all subsequent session traffic to an arbitrary internal endpoint by injecting unvalidated `directConnectHost` and `directConnectPort` parameters when `directConnect(true)` is enabled, potentially leading to session traffic interception, network pivoting, and cloud credential theft.