The autoloan npm package contains malicious code that executes a beacon script during installation and module import. This script collects system identifiers and sends them to a hardcoded external server over HTTPS with disabled TLS verification. The package itself provides no real functionality and appears designed for install-time tracking or dependency confusion attacks.
The npm package @c1-ease/core contains malicious code that collects host identifiers during installation or import and sends them to a hardcoded external server. The package exports a no-op Proxy with no functional API, and its main observable effect is the transmission of installer environment data to a suspicious endpoint with disabled TLS verification. This behavior is consistent with dependency confusion or typo-squatting attacks.
The npm package @c1-ease/autoloan contains malicious code that executes a postinstall script and triggers on import. It sends system information to a hardcoded external server over HTTPS with disabled certificate validation, enabling data exfiltration. The package uses dependency-confusion naming to masquerade as an internal namespace, causing silent execution in build and CI environments.
The @c1-ease/bank npm package contains malicious code that executes a postinstall script and runtime beacon sending host system information to an attacker-controlled server with disabled TLS verification. The package provides no legitimate functionality and appears designed for dependency confusion or install tracking, exfiltrating sensitive environment data upon installation or require.
The npm package @c1-ease/card is a malicious dependency confusion shim that exports a no-op Proxy and includes a postinstall and require-time beacon. This beacon collects host identifiers and sends them unprotected to a hardcoded third-party server with disabled TLS validation, enabling data exfiltration. The package has no legitimate functionality and is designed to be a stealthy exfiltration vector in supply chain attacks.
The npm package @c1-ease/commerce versions 2.0.0 and 2.0.1 contains malicious code that executes a post-install script and a runtime callback beacon. This beacon sends host identifiers such as hostname, install path, current working directory, and Node.js version to a suspicious external server with disabled TLS verification. The package acts as a compatibility shim with no legitimate functionality, and the beacon fires automatically on install and import without user consent. This behavior is consistent with a dependency confusion or namespace squatting attempt targeting internal @c1-ease packages.
A critical security update for PHP 8.2 on Red Hat Enterprise Linux addresses multiple vulnerabilities including remote code execution, local privilege escalation, request smuggling, cross-site scripting, and denial of service. The update rebases PHP to version 8.2.36 and fixes issues such as use-after-free bugs, out-of-bounds reads/writes, and signature bypasses. These vulnerabilities affect PHP when used in various configurations including PHP-CGI and PHP-FPM. The advisory provides fixes for all identified issues and recommends applying the update to mitigate these risks.
CVE-2026-134572 is a SQL injection vulnerability in the itsourcecode Insurance Management System version 1.0. The flaw exists in the /deleteAgent.php file where the agent_id parameter can be manipulated to execute unauthorized SQL commands. This vulnerability allows remote attackers to exploit the system without user interaction. Public exploit code is available for this issue.
Mailflare versions before commit 48957fb contain a host header injection vulnerability in the password reset request handler. This flaw allows unauthenticated attackers to manipulate the Host header when APP_URL is unset, poisoning password reset links. By submitting a reset request with a malicious Host header for a victim's email, attackers can capture the single-use reset token when the victim clicks the link, enabling account takeover.
Open5GS AMF versions up to 2.8.0 have a vulnerability where resources are not properly released after their effective lifetime. Authenticated user equipment (UE) can exploit this by triggering failures in Create SM Context response processing, causing session orphaning and degradation of SMF availability.
OpenAirInterface CN5G SMF versions up to 2.2.2 have a vulnerability where resources are not properly released after PDU session release requests. This flaw allows authenticated user equipment (UE) to repeatedly establish and release PDU sessions, causing stale SM policy associations to accumulate at the PCF and potentially exhaust its memory.
CVE-2026-134657 is a memory leak vulnerability in OpenAirInterface CN5G PCF versions up to 2.2.2. The flaw occurs in the pcf_smpc::delete_sm_policy_handler() function, which fails to release memory for stale UE IPv4, SUPI, and DNN index entries after association deletion. An attacker can exploit this by repeatedly creating and deleting SM policies via N7 interface PDU session establishment and release, causing unbounded memory growth and potential disruption of association lookups.
free5GC SMF versions up to 1.4.6 have a vulnerability where resources are not properly released after failed PDU session establishments, causing orphaned charging sessions on the CHF. This can degrade the availability of the CHF by accumulating unused sessions.
OpenAirInterface CN5G AMF versions before 2.2.2 have a vulnerability where resources are not properly released after failed NRF lookup or SMF selection. This causes stale PDU session contexts to remain, leading to deregistration delays and stalling of NAS processing for other authenticated UEs requesting PDU sessions when no SMF is selectable.
GPAC MP4Box versions up to 26.07.0 contain a type confusion vulnerability in the xmt_parse_command() function. This flaw arises from casting target nodes to GF_ParentNode without proper validation, allowing crafted XMT-A files to cause invalid memory reads and crash the application.
GPAC MP4Box versions up to and including 26.07.0 contain a NULL pointer dereference vulnerability in the function gf_odf_encode_ui_config() located in src/odf/descriptors.c. This flaw can be triggered by processing a crafted XMT file that declares an interaction stream without a UIConfig descriptor, leading to a denial of service via process crash.
GPAC MP4Box versions up to 26.07.0 contain a double free vulnerability in the XMT-A scene loader. This flaw occurs when processing a crafted XMT file with malformed nesting, leading to heap corruption and potential process crash.
GPAC MP4Box versions up to 26.07.0 contain a use-after-free vulnerability in the XMT scene loader. This flaw occurs because the loader retains a pointer to a PROTO sub-graph that has been freed, allowing a crafted XMT file to cause a crash by updating a freed graph's referencing_commands list.
GPAC MP4Box versions up to 26.07.0 contain a NULL pointer dereference vulnerability in the compositor component. This flaw occurs when processing crafted XMT scenes that cause the compositor to dereference an uninitialized sensor node pointer, leading to a crash and denial of service.