ExploitBench: A Capability Ladder Benchmark for LLM Cybersecurity Agents

Exploitation is not a binary event. It is a ladder of acquiring progressive capabilities, from executing a single buggy line of code to taking full control of the target. However, existing LLM security benchmarks treat a crash as exploitation success. That single binary outcome collapses the hard...

AI Native Asset Intelligence

Modern security environments generate fragmented signals across cloud resources, identities, configurations, and third-party security tools. Although AI-native security assistants improve access to this data, they remain largely reactive: users must ask the right questions and interpret...

Stable Agentic Control: Tool-Mediated LLM Architecture for Autonomous Cyber Defense

Agentic systems involved in high-stake decision-making under adversarial pressure need formal guarantees not offered by existing approaches. Motivated by the operational needs of security operations centers SOCs that must configure endpoint detection and response EDR policies under adversarial...

Self-Adaptive Multi-Agent LLM-Based Security Pattern Selection for IoT Systems

The adoption of Internet of Things IoT systems at the network edge of smart architectures is increasing rapidly, intensifying the need for security mechanisms that are both adaptive and resource-efficient. In such environments, runtime defence mechanisms are no longer limited to detection alone b...

SNF - Shadow Network Fingerprinting Engine

SNF Shadow Network Fingerprinting Engine is a 100% offline, air-gap-native passive network intelligence engine written entirely in Rust. It was designed from the ground up for environments where outbound connectivity is not just unavailable but prohibited: classified defense networks, nuclear...

SUSE-SU-2026:21374-1 Security update for erlang

This update for erlang fixes the following issues: Security issues fixed: - CVE-2026-21620: improper isolation and compartmentalization can lead to TFTP relative path traversal and remote arbitrary reads/writes bsc1258663. - CVE-2026-23941: improper handling of duplicate Content-Length headers in...

CLSA-2026-1776443255 libxslt: Fix of CVE-2023-40403

CVE-2023-40403: make generate-id deterministic to prevent memory layout leak...

Deterministic + Agentic AI: The Architecture Exposure Validation Requires

Few technologies have moved from experimentation to boardroom mandate as quickly as AI. Across industries, leadership teams have embraced its broader potential, and boards, investors, and executives are already pushing organizations to adopt it across operational and security functions. Pentera’s...

EUVD-2025-209419

Nitro PDF Pro for Windows 14.41.1.4 contains a heap use-after-free vulnerability in the implementation of the JavaScript method this.mailDoc. During execution, an internal XID object is allocated and then freed prematurely, after which the freed pointer is still passed into UI and logging helper...

skilleton has improper input handling in repository/path processing

Summary skilleton versions prior to 0.3.1 include security-related weaknesses in repository normalization and path handling logic. Version 0.3.1 contains fixes and additional test coverage for these issues. Affected Versions =0.3.1 Impact In affected versions, crafted input could trigger unsafe o...

GHSA-5G3J-89FR-R2VP skilleton has improper input handling in repository/path processing

Summary skilleton versions prior to 0.3.1 include security-related weaknesses in repository normalization and path handling logic. Version 0.3.1 contains fixes and additional test coverage for these issues. Affected Versions =0.3.1 Impact In affected versions, crafted input could trigger unsafe o...

dst-engine

DST: Deterministic Security Testing Static analysis that does...

CVE-2026-22665

prompts.chat prior to commit 1464475, contains an identity confusion vulnerability due to inconsistent case-sensitive and case-insensitive handling of usernames across write and read paths, allowing attackers to create case-variant usernames that bypass uniqueness checks. Attackers can exploit...

CVE-2026-33184 nimiq/core-rs-albatross: Discovery handshake limit could underflow and later provoke a deterministic overflow panic

nimiq/core-rs-albatross is a Rust implementation of the Nimiq Proof-of-Stake protocol based on the Albatross consensus algorithm. Prior to version 1.3.0, the discovery handler accepts a peer-controlled limit during handshake and stores it unchanged. The immediate HandshakeAck path then honors lim...

CVE-2026-33184

The CVE affects the nimiq/core-rs-albatross Rust implementation of the Nimiq PoS Albatross protocol. Before 1.3.0, the discovery handshake allowed a peer-controlled limit to be stored, causing the HandshakeAck path to honor limit=0 and return zero contacts. After establishment, the periodic updat...

CVE-2026-23407

A flaw was found in AppArmor, a security module within the Linux kernel. A local user could exploit this vulnerability by providing a specially crafted Deterministic Finite Automaton DFA, a set of rules for pattern matching, to the verifydfa function. This malformed input causes the system to...

CVE-2026-23406

In the Linux kernel, the following vulnerability has been resolved: apparmor: fix side-effect bug in matchchar macro usage The matchchar macro evaluates its character parameter multiple times when traversing differential encoding chains. When invoked with str++, the string pointer advances on eac...

CVE-2026-23406

In the Linux kernel, the following vulnerability has been resolved: apparmor: fix side-effect bug in matchchar macro usage The matchchar macro evaluates its character parameter multiple times when traversing differential encoding chains. When invoked with str++, the string pointer advances on eac...

CVE-2025-67114

Use of a deterministic credential generation algorithm in /ftl/bin/calcf2 in Small Cell Sercomm SCE4255W FreedomFi Englewood firmware before DG3934v3@2308041842 allows remote attackers to derive valid administrative/root credentials from the device's MAC address, enabling authentication bypass an...

SUSE CVE-2026-25645

Requests is a HTTP library. Prior to version 2.33.0, the requests.utils.extractzippedpaths utility function uses a predictable filename when extracting files from zip archives into the system temporary directory. If the target file already exists, it is reused without validation. A local attacker...