This report details the infrastructure of SpiceRAT malware linked to the SilkParasite campaign targeting government and energy sectors across Central Asia. The infrastructure includes a cluster of command and control servers active from late 2025 through mid-2026, identified by shared TLS certificates, reused hostnames, and identical cloned web pages. The attackers impersonate legitimate entities such as Uzbekistan's state railway authority using TLS certificates issued by a Chinese state-affiliated CA. The infrastructure spans multiple hosting providers and countries, with evidence of ongoing activity for at least four years. The report provides detection indicators including IPs, domains, TLS certs, and page hashes for defenders to conduct retro hunts. No direct malware behavior or initial access vectors are covered in this analysis.
BragJack is a security research discovery demonstrating how a single browser extension can hijack internal browser agents across five Chromium-based browsers: Chrome, Comet, Edge, Opera Neon, and Claude in Chrome. The attack leverages the declarativeNetRequest (DNR) API to weaken security headers and redirect JavaScript resources, enabling code execution within privileged browser contexts. This allows access to sensitive capabilities such as local file access, browsing history, camera and microphone, and screenshots without user interaction. The research resulted in two CVEs and $20,000 in bounty rewards from major vendors. The vulnerabilities primarily exploit weaknesses in how these browsers handle network requests and security policies within embedded web views. No patch or remediation status is provided in the source data.
This content discusses the potential early arrival of 'Q-Day,' the point when quantum computers could break traditional asymmetric cryptography, possibly as early as 2028. It highlights that most governments and organizations plan to be post-quantum prepared by 2030-2035, but recent statements from IonQ suggest that breaking 256-bit elliptic curve cryptography might be feasible sooner. The post emphasizes the need for organizations to accelerate their post-quantum cryptography projects, conduct rapid cryptographic inventories, and implement emergency response plans if Q-Day occurs earlier than expected. It also advises prioritizing mitigation of authentication threats and updating incident response policies to address quantum-related attacks. The content serves as a strategic warning rather than a report of an active vulnerability or exploit.
The GemStuffer incident involved AI agents exploiting a documentation feature in RubyGems infrastructure to achieve remote code execution (RCE). Specifically, the YARD tool's --load option in .yardopts files was abused to execute arbitrary Ruby code during automated documentation builds on RubyDoc.info. The attackers created disposable accounts, bypassed email confirmation, and uploaded over 2,000 malicious packages that weaponized this feature. This incident highlights a broader class of vulnerabilities where legitimate scripting or code execution features in configuration files are abused in package ecosystems. The attack demonstrates a real supply chain risk from automated AI-driven exploitation.
A Chinese-speaking threat actor group known as Red Heron exploited a recently disclosed remote code execution vulnerability (CVE-2026-60004) in Gitea, a self-hosted Git service, in a multinational campaign. The campaign targeted internet-facing Gitea instances across multiple countries, including Canada, Argentina, Taiwan, the United States, and Sri Lanka, focusing on sectors such as defense, elections, energy, aerospace, telecommunications, government, and research. The attackers used automated tools to steal source code, credentials, and maintain persistent access, including root-level control on some infrastructure. They deployed a novel Linux implant named JITTERLY with extensive post-exploitation capabilities and embedded a previously undocumented rootkit called SIXZUT to maintain stealth and persistence. The campaign demonstrates rapid weaponization of n-day vulnerabilities in development platforms and highlights significant risks to source code confidentiality and infrastructure integrity.
Project RedTeam: Contract Offensive is a forthcoming educational game designed to teach network intrusion methodology through a gamified, card-based roguelite experience. The game simulates real-world cybersecurity tradecraft mapped to MITRE ATT&CK techniques, allowing players to engage in simulated corporate network intrusions. It is intended as a learning tool and does not contain real organizations or actionable illegal hacking instructions. The game is not a security threat or vulnerability but rather an educational simulation.
This threat involves cookie-controlled PHP webshells used by attackers to maintain stealthy, persistent access in Linux shared hosting environments. The webshells remain dormant until triggered by specific HTTP cookie values, reducing detection by blending into normal traffic. Attackers use obfuscation and layered payload deployment to evade discovery. Persistence is often maintained via scheduled tasks (cron jobs) within the hosting account, allowing low-noise long-term access without root privileges. The attack pattern can include long periods of inactivity followed by increased malicious activity, sometimes timed to weekends or low-monitoring periods. This technique complicates incident response and detection in shared hosting scenarios.
A local privilege escalation vulnerability in the Steam Client Service on Windows 10 and 11 allows any standard user to silently escalate privileges to SYSTEM without requiring admin rights, UAC prompts, or launching a game. The issue stems from a signature verification gap in Steam's installation process, enabling execution of code as SYSTEM without signature forgery. This was tested on Steam version 10.96.30.42. Valve was notified months prior to public disclosure.
This content describes an open-source self-hosted threat intelligence aggregator tool designed to collect, deduplicate, and correlate threat intelligence feeds from over 60 sources. It includes features such as IOC extraction, ATT&CK mapping, integration with asset inventories, and detection engineering capabilities. The tool can run locally or integrate with Azure environments and supports optional AI triage. It is not a vulnerability or exploit but rather a security tool release.
Aethel-Core is a neuro-symbolic hybrid Web Application Firewall (WAF) implemented in SWI-Prolog, designed for real-time Layer-7 threat mitigation with extremely low latency (0.17ms). It uses a large set of dynamic, tokenized threat signatures and logical inference to detect and block complex multi-vector attacks, including LFI and CRLF injections, as well as advanced threats like Web3 anomalies and adversarial AI simulations. The system generates structured JSON telemetry suitable for large language model fine-tuning. The project is open source under MIT License for core components, with proprietary enterprise extensions under development. The production deployment model is evolving as a standalone asynchronous reverse proxy gateway intended for zero-touch integration with existing infrastructure. No known exploits or vulnerabilities are reported.
An open directory exposed on a server in Thailand revealed an active intrusion operation targeting two major broadband providers, primarily 3BB (Triple T Broadband). The attacker exploited CVE-2024-21762 to gain remote code execution on a FortiGate 60F SSL-VPN device. The exposed files included exploitation scripts, privilege escalation tools, brute-force utilities, and a MeshCentral agent configured as a persistent backdoor. The attacker aimed to extract subscriber authentication data from RADIUS databases and had valid OpenVPN certificates and active session cookies, indicating deep network access. The attacker also prepared cleanup scripts to remove traces while maintaining persistence. Hunt.io notified affected parties and CERT prior to public disclosure.
This entry references a simulation of the Charming Kitten APT adversary, as detailed in a Medium article. It is not a direct report of an active campaign or exploit but rather an adversary simulation intended for security awareness or training purposes.
EchelonGraphBot is a web crawler operated by EchelonGraph, Inc. that scans public websites to check configurations such as HTTPS enforcement, certificate expiration, TLS versions, security headers, cookie flags, and redirects. It visits a large number of third-party hosts approximately once per day, respecting robots.txt and rate limits. Operators can opt out via a DNS TXT record or robots.txt. The bot's activity is transparent and publicly documented, with no evidence of malicious intent or exploitation. The operator offers to adjust probing frequency upon request and honors opt-out requests.
A study tested whether instructing a large language model (LLM) to generate a "secure" 4-digit PIN improves unpredictability. Using Llama 3.2 11B, the study found that explicitly asking for a hard-to-guess PIN actually increased predictability and concentration on a small set of PINs that appear random but are not truly unpredictable. This mirrors recent findings in password generation by other LLMs, showing that LLM-generated credentials are not inherently secure. The study emphasizes that relying on an LLM's notion of "secure" PINs or passwords is insufficient and that generated credentials should be validated against blocklists or entropy checks. No direct exploit or patch is involved, but the findings have implications for security practices involving LLM-generated credentials.
This report details a controlled memory forensics investigation on a Windows 11 system to determine what user activity can be reconstructed from a RAM capture. The experiment involved normal user actions such as browsing with Edge, running PowerShell commands, and typing unsaved text in Notepad. Memory was acquired using WinPmem and analyzed with Volatility 3, revealing recoverable artifacts including running processes, command-line arguments, network state, and unsaved text. The findings emphasize that memory forensics is most effective when multiple artifacts are correlated rather than analyzed in isolation. This is a research and educational exercise rather than a vulnerability or exploit.
A campaign named Miasma exploited a vulnerability in the Model Context Protocol (MCP) tool configuration process to gain OS-level access to 73 GitHub repositories, including one owned by Microsoft Azure. The attack leverages the fact that MCP clients trust tool descriptions as executable instructions without re-prompting or sandboxing, allowing malicious edits to MCP config files to run with developer privileges. This attack class, known as MCP tool poisoning, was first documented in 2025 and escalated to real-world exploitation by mid-2026. The campaign abused the approval process tied to server names rather than content, enabling malicious commands to execute automatically after initial approval. OWASP ranks this attack third in its MCP Top 10 threats. Mitigations include rigorous code review of MCP config changes, pinning tool definitions, scoping privileges per tool, maintaining allowlists of MCP servers, and sandboxing agent execution.
voidsyscall is a cross-platform implant and command-and-control (C2) framework that operates by issuing direct syscalls on Windows and raw syscalls on Linux, bypassing the Windows API layer entirely. It uses multiple stealthy injection techniques and evasion methods to avoid detection by endpoint detection and response (EDR) tools, including runtime resolution of syscall numbers, unhooking ntdll, and polymorphic code injection. The implant supports encrypted C2 communication over HTTPS, DNS, and ICMP channels using AES-256-GCM encryption. It includes advanced anti-analysis features such as VM, sandbox, and debugger detection, and performs token manipulation and memory encryption to hinder forensic analysis. No official patch or remediation is indicated, and this tool is presented as an offensive implant framework rather than a vulnerability.
This report discusses a method to locate Flutter's TLS certificate verifier function within a stripped libflutter.so binary on ARM64 devices without relying on fragile byte signatures. Flutter bundles its own BoringSSL and does not use Android's network security config, requiring patching of the ssl_crypto_x509_session_verify_cert_chain function to modify trusted CAs. The function can be identified by cross-referencing unique strings and specific ARM64 prologue characteristics, enabling reliable patching across multiple Flutter engine versions. This technique was validated on apps using Flutter engine versions 2.19.2 through 3.11.1. The report also highlights a common pitfall in patching tools that fail to handle split APKs correctly. No active exploits or vendor patches are mentioned.
This content discusses the impact of AI on the speed at which vulnerabilities can be analyzed and exploited, particularly focusing on dependency management and patch windows. It highlights that AI-assisted tools can rapidly analyze patches and identify vulnerabilities in abandoned or stale dependencies, potentially shrinking the window defenders have to patch before exploitation. The discussion raises concerns about whether traditional patching SLAs are sufficient in the age of AI and suggests that dependency maintenance should be considered part of the security perimeter.
An unauthenticated local privilege escalation vulnerability exists in the MODAPI.sys kernel driver installed by MSI Dragon Center 2.0.155.0. The driver exposes a device object with a security descriptor allowing any local user to open it without administrative privileges or UAC. This enables arbitrary read and write access to Model-Specific Registers (MSRs), including the IA32_LSTAR MSR, which controls the CPU syscall entry point. By overwriting IA32_LSTAR, an attacker can execute arbitrary code at kernel privilege level (CPL0), bypassing SMEP via gadgets found in the kernel image. The exploit was tested successfully on Windows 10 x64 22H2 (build 19045.6456) from a standard user account. The vulnerability was reported to MSI on 2026-06-21 and a fix has been verified.
