A missing privilege check in older Arcane versions lets holders of a specific delegation permission reset the global administrator password and seize full control.
Attackers can inject malicious objects into the LearnPress plugin via checkout parameters, gaining critical access to WordPress sites without logging in.
A critical use-after-free flaw in MariaDB enables remote code execution when combined with another vulnerability. Affected servers should apply the vendor patch.
Treating firmware updates as simple patches ignores the deep hardware trust chain, allowing attackers to persist even after a full system wipe or network isolation.
A single misconfigured container can grant attackers full access to the underlying host system, bypassing application-level security controls entirely.
Supply chain attacks hide inside trusted code dependencies, exploiting the trust you place in automated installers rather than your application logic directly.
Carriage return and line feed characters let attackers split HTTP responses to inject new headers, enabling cache poisoning and session hijacking without executing code.
NetScaler often fails silently, so you must monitor for subtle protocol deviations and unexpected certificate changes rather than waiting for obvious service outages.
Misconfigured web servers expose version control metadata, allowing attackers to reconstruct entire repositories and steal secrets without exploiting software bugs.
Small businesses face disproportionate risk from Exchange flaws because attackers exploit the gap between public internet exposure and limited internal patching capacity.
Kernel exploits bypass application boundaries by manipulating memory structures, granting attackers full system control that survives reboots and ignores file permissions.
Runtime application self-protection fails when teams treat it as a silver bullet rather than a context-aware monitoring layer that requires careful tuning.
Canonical host validation at the application layer is the only reliable defense against host header injection, as network-level filters cannot distinguish legitimate multi-tenant traffic from malicious overrides.
Attackers can crash Rust-based AWS SDK servers by sending small, deeply nested data requests that exhaust stack space without needing credentials.
Java deserialization flaws allow attackers to execute arbitrary code by tricking applications into reconstructing malicious object states from untrusted data.
A critical Linux kernel vulnerability allows remote attackers to corrupt network headers by sending packets with invalid checksum metadata, potentially causing system mischief.
An authorization bypass in the low-code platform allows authenticated users to access and execute queries belonging to other organizations on shared instances.
A command injection flaw in older Komodo agents allows authenticated users to execute arbitrary root commands on servers, risking full infrastructure compromise.
A flaw in the AWS HealthLake MCP server lets remote users with valid logins trick the system into sending temporary security keys to attacker-controlled servers.
A server-side request forgery in an AI agent tool allowed attackers to steal operator credentials via crafted prompts.
A reflected XSS in Siemens Teamcenter allows zero-interaction code execution, enabling attackers to alter account permissions without user clicks.
A critical flaw in Hitachi Vantara Remot Ops Agent allows attackers to bypass login by signing their own access tokens.
Unauthenticated users can hijack accounts in Better Auth versions prior to 1.7.7 by misusing OAuth state values as verification tokens, a critical flaw rated 9.3 by NVD.
Unauthenticated attackers can execute commands as root on Cisco Secure Email Gateway appliances by sending a single malicious email with an attachment.