Daily Edition · Page 5

THE DAILY INDEX

Technology, AI, software, security, crypto and business

ATTACKS & VULNERABILITIES

TLDR InfoSec

CRITICAL INFRA HONEYWELL CCTVS VULNERABLE TO AUTH BYPASS FLAW

CISA warned of CVE-2026-1670 (CVSS 9.8), a missing authentication flaw in multiple Honeywell CCTV models that allows unauthenticated attackers to remotely change the password recovery email address, enabling full account takeover and unauthorized access to camera feeds. Affected models include the I-HIB2PI-UL and several SMB NDAA-compliant PTZ cameras deployed in commercial and critical infrastructure environments. Organizations should isolate these devices behind firewalls, restrict network exposure, and contact Honeywell for patch guidance.

RED VULNS RISING: EXAMINING CHINESE NATIONAL VULNERABILITY DATABASES

Bitsight analyzed China's two national vulnerability databases—CNNVD, overseen by the Ministry of State Security, and CNVD, operated by CNCERT—and found that while they largely mirror CVE publication, approximately 1,400 entries were published in Chinese databases before becoming public in CVE, often by several months. China's 2021 RMSV regulations tightened domestic vulnerability disclosure controls, including mandatory 48-hour reporting to the government and prohibitions on sharing PoC exploits. There was a notable decline in non-CVE vulnerability publications from CNVD after the regulation, although CNNVD has recently seen a resurgence. Security leaders should account for potential blind spots in Western vulnerability ecosystems, as some Chinese database entries lack CVE equivalents and may represent vulnerabilities unknown to Western defenders.

CLINEJECTION — COMPROMISING CLINE'S PRODUCTION RELEASES JUST BY PROMPTING AN ISSUE TRIAGER

A prompt-injected GitHub issue title could drive Cline's Claude-based triage bot to execute arbitrary commands in CI, then use GitHub Actions cache poisoning to hijack nightly build workflows and steal VS Code Marketplace, OpenVSX, and npm publishing tokens. This chain could have enabled a massive supply-chain attack on millions of developers via malicious extension or CLI updates.

DECRYPTING MULTIDESK PASSWORDS

MultiDesk RDP client's password encryption was reverse-engineered, revealing RC4 encryption with per-user keys stored in the HKEY_CURRENT_USER\Software\MultiDesk\key registry hive across both legacy (v3.16) and modern (v14.0) versions. Version 3.16 used base64-encoded RC4 ciphertext with rdtsc-generated keys, while version 5+ introduced a salted key derivation scheme using CryptGenRandom, but remained fundamentally RC4-based. Pentesters should add MultiDesk.xml, MultiDesk.multidesk, and the associated registry keys to their internal recon checklists, as credential recovery is trivial with local admin or user-level access to the target machine.

CITIZEN LAB FINDS CELLEBRITE TOOL USED ON KENYAN ACTIVIST'S PHONE IN POLICE CUSTODY

Citizen Lab confirmed with high confidence that Cellebrite forensic extraction tools were used on Kenyan pro-democracy activist Boniface Mwangi's Samsung phone while in police custody following his July 2025 arrest, potentially enabling full data extraction including messages, passwords, and financial information. The findings follow a similar report of Cellebrite misuse in Jordan and coincide with Amnesty International's discovery of Intellexa's Predator spyware targeting an Angolan journalist via WhatsApp, marking the first confirmed Predator case against civil society in Angola. The cases highlight the growing global abuse of commercial surveillance tools against activists and journalists, with Predator's sophisticated anti-forensics capabilities including crash monitoring, recording indicator suppression, and granular operator control over failed deployments.

KEEPING GOOGLE PLAY & ANDROID APP ECOSYSTEMS SAFE IN 2025

Google reported blocking over 1.75 million policy-violating apps and banning 80,000 malicious developer accounts from Google Play in 2025, while Play Protect's real-time scanning identified more than 27 million new malicious apps from outside the store across 350 billion daily scans. Enhanced fraud protection expanded to 185 markets covering 2.8 billion devices, blocking 266 million risky sideloaded installation attempts. New defenses included in-call scam protection preventing users from being socially engineered into disabling Play Protect, and Android 16's one-line-of-code tapjacking protection for sensitive app screens.

TLDR IndexPage 5