Author

Saugata Datta

I’m an IT/DevOps professional with many years of hands-on experience across Windows, Linux, Automation, Backup and Cloud. I love building real-world automations to make infrastructure reliable, secure, and self-healing. Through this blog, I share practical learnings from enterprise projects and my home-lab experiments on Cloudflare, Self-Hosting and everything in between. Always tinkering, always improving.
Page 14
Chainalysis Confirms Customer Data Exposure After Attackers Used “Offensive Agentic Techniques”

Chainalysis Confirms Customer Data Exposure After Attackers Used “Offensive Agentic Techniques”

Blockchain analytics firm Chainalysis has disclosed a security incident after attackers used what the company described as "offensive agentic techniques" to gain unauthorized access to internal systems and tools. The company says some customer information—names, email addresses, user IDs, and organizational role labels—was accessed, but it has found no evidence that sensitive investigation data, authentication credentials, or session tokens were exposed. Chainalysis is still determining the full scope of accessed…

Saugata Datta3 min read177
Reddit’s move to make full content available only to logged-in users

Reddit’s move to make full content available only to logged-in users

For two decades, Reddit allowed most of its content to be browsed without an account. That era is changing: co‑founder Steve Huffman announced the company will gradually shift toward a “logged‑in service,” limiting some content for visitors who aren’t signed in. The stated goals are straightforward—deliver better personalization for members and reduce unauthorized data harvesting that has become a persistent problem for the platform. Why Reddit is changing access Huffman…

Saugata Datta3 min read177
Cursor: Malicious Claude Code Extension in VS Code Marketplace Carried SSH Backdoor for 11 Hours

Cursor: Malicious Claude Code Extension in VS Code Marketplace Carried SSH Backdoor for 11 Hours

Cursor confirmed that weaponized builds of the Claude Code extension briefly appeared in its Visual Studio Code extension marketplace on October 29, 2026. For roughly eleven hours those tampered packages attempted to harvest host details and install an SSH backdoor intended to provide persistent remote access. Cursor says the compromised builds were not produced by Anthropic and that distribution was limited to specific versions of the extension. Timeline and affected…

Saugata Datta3 min read188
Chainalysis Confirms Agentic Cyberattack That Exposed Customer Account Details

Chainalysis Confirms Agentic Cyberattack That Exposed Customer Account Details

Chainalysis has confirmed a cyberattack that used what the company described as “offensive agentic techniques” to access internal systems and tools, exposing customer account information. The firm says the incident was detected in October and that it has notified affected customers while engaging outside cybersecurity experts to investigate and harden its platform. What Chainalysis says about the incident Chainalysis reported that an external attacker accessed user IDs, full names, email…

Saugata Datta3 min read182
TeamViewer Bug Lets Remote Hosts Crash or Run Code Against Technicians

TeamViewer Bug Lets Remote Hosts Crash or Run Code Against Technicians

A recently disclosed vulnerability in TeamViewer can turn an ordinary remote support session into a serious risk for the technician on the other end. Tracked as CVE-2026-111585 and rated High with a CVSS score of 8.8, the flaw lets a malicious host crash the viewer application and, in some cases, execute code on the technician’s machine after a connection has been established. How the flaw works TeamViewer separates sessions into…

Saugata Datta3 min read192
Inside the 60,471 Attempts: What the WordPress 7.1.4 Patch Actually Stopped

Inside the 60,471 Attempts: What the WordPress 7.1.4 Patch Actually Stopped

Hackers flooded WordPress sites with tens of thousands of requests targeting a Core remote code execution (RCE) weakness in the hours after a security patch was released. Wordfence’s threat intelligence recorded 60,471 blocked attempts in a 24‑hour window — a clear signal that attackers were actively probing the update process, even though the majority of those requests were firewall‑blocked and not confirmed compromises. What attackers were trying to exploit The…

Saugata Datta3 min read183
Cloudflare’s Q3 Surge: 18.2 Million Network-Layer DDoS Attacks and Changing Threat Patterns

Cloudflare’s Q3 Surge: 18.2 Million Network-Layer DDoS Attacks and Changing Threat Patterns

Threat actors ramped up automated disruption in the third quarter of 2026, with Cloudflare reporting a dramatic rise in the frequency and variety of distributed denial-of-service (DDoS) attacks. The company mitigated 18.2 million network-layer (L3) DDoS attacks and handled 14.4 trillion HTTP DDoS requests during the quarter — a scale that underscores how persistent and rapidly evolving DDoS activity has become. Quarterly volume and pace Cloudflare’s numbers translate to roughly…

Saugata Datta2 min read202
Apple’s New Wave of Spyware Warnings: What iPhone Users Need to Know

Apple’s New Wave of Spyware Warnings: What iPhone Users Need to Know

Apple has sent a new round of targeted threat notifications to selected iPhone users, warning they may have been the focus of sophisticated mercenary spyware campaigns. Announced in a notice shared by security researcher Costin Raiu on October 30, 2026, this latest wave is separate from the alerts Apple sent in August and is being delivered to users in 99 countries. Apple says it has now notified users in over…

Saugata Datta4 min read205
Anthropic Exploit Reveals Critical Linux Kernel Flaw (CVE-2026-98432) That Grants Remote Root on MPTCP Servers

Anthropic Exploit Reveals Critical Linux Kernel Flaw (CVE-2026-98432) That Grants Remote Root on MPTCP Servers

Security researchers at Anthropic have developed an exploit for a critical Linux kernel vulnerability—tracked as CVE-2026-98432—that can allow unauthenticated attackers to execute commands as root on servers running Multipath TCP (MPTCP) listeners. The flaw is a use-after-free in the kernel's IPv4 TCP networking code and has been scored 9.8/10 by the Linux kernel CVE team, although some vendors assess exploitation complexity differently. What the flaw is CVE-2026-98432 stems from a…

Saugata Datta3 min read202
Hardcoded Database Password in Applebee’s App Exposed Customer Records

Hardcoded Database Password in Applebee’s App Exposed Customer Records

Applebee’s patched a security flaw after a researcher found a database password embedded inside the chain’s Android app, a lapse that allowed access to customers’ personal information. TechCrunch held the story until the issue was confirmed fixed. The exposed credential, discovered in the app’s source code, could be extracted by anyone with the app and used to access the company’s customer database. Discovery and verification Robert Wiggins, a U.K.-based cybersecurity…

Saugata Datta3 min read203
Active exploitation in Drupal Commerce Core (CVE-2026-118539): urgent patching for online stores

Active exploitation in Drupal Commerce Core (CVE-2026-118539): urgent patching for online stores

Drupal has issued a public advisory warning that a critical flaw in the Commerce Core module is being actively exploited, placing online stores at risk of unauthorized changes to orders and other store data. Published on October 28, 2026, the advisory identifies missing access checks in routes and entity fields that could let attackers reach functions or modify data they should not control. Drupal assigned the issue a high risk…

Saugata Datta4 min read195
Aur0ra’s AI-Driven Leap: How Cursor Agents Helped Hackers Reach Domain Admin

Aur0ra’s AI-Driven Leap: How Cursor Agents Helped Hackers Reach Domain Admin

Ransomware operators tied to the Aur0ra (Aur0ra/Aur0ra) family used AI coding agents to shorten the path from stolen credentials to full domain administrator control, then collected data and prepared ransomware deployment. Sophos researchers traced the activity to a group tracked as GOLD OGRE, which combined human-provided access and goals with Cursor AI agents that automated many intrusion steps. The result: lower skill required to run complex intrusions and faster, more…

Saugata Datta3 min read193
🧰

Free Tools: IP, DNS, SSL

40+ free network and other tools in one place. No signup. Click explore or visit https://technochat.in/tools

Explore