Blog: AI, Cybersecurity, Linux & DevOps

Technical watch, field-tested experience and practical guides for sysadmins and DevOps who care about robustness.

Nouveau : le guide complet de l'IA locale en 2026 — matériel, modèles, déploiement et coût réel, en un seul point d'entrée. Lire le guide →

Recent articles

79 articles

AI 8 min read

Local RAG with Ollama: an assistant that reads YOUR documents, 100% offline

Build a privacy-first RAG assistant on your own documents: embeddings, vector DB, chunking and a local LLM. With qwen3-embedding hitting 70.58 on multilingual MTEB, local finally rivals commercial APIs.

AI 9 min read

Local LLM Runtimes in 2026: llama.cpp, Ollama, vLLM, LM Studio, TGI, Which One to Pick?

An honest comparison of local LLM inference engines in 2026: vLLM hits ~793 tok/s under concurrent load versus ~41 for Ollama, yet at a single user the gap drops below 10%. When to use each.

AI 9 min read

Fine-tune an LLM locally with LoRA and QLoRA: VRAM, datasets and realistic expectations

How much VRAM do you really need? QLoRA fine-tunes a 7B on 8-10 GB, a 13B on a 24 GB GPU. We debunk the myths: when fine-tuning beats RAG, and when it falls flat.

AI 7 min read

Local AI vs Cloud API: How Many Tokens Until Your GPU Actually Pays Off?

The honest break-even math between a 2,600 EUR RTX 5090 and the Claude/GPT API in 2026: hardware amortization, electricity, tokens/month. Spoiler: you need to target 50 to 100 million tokens per month.

AI 8 min read

NPU, TOPS and AI: what these neural chips really do (and why TOPS lie about LLMs)

AMD XDNA 2, Apple Neural Engine, Intel NPU 5: these chips advertise 38 to 55 TOPS, yet a 7B LLM generates text at only ~7 tok/s on them. A breakdown of what an NPU is actually good for, and what belongs on the GPU.

AI 8 min read

100% Local Audio Transcription: Self-Hosting Whisper and faster-whisper

Whisper running locally, no cloud: models, GPU/CPU performance, real-time vs batch, accuracy (WER) and privacy. With faster-whisper, large-v3-turbo fits in 1.5 GB of VRAM at INT8.

Security 8 min read

Securing an Exposed Ollama Instance: The Real Risks of Local AI on a Network

Ollama ships with no authentication by default. Over 300,000 instances are reportedly exposed on the internet in 2026. Model theft, RCE, LLMjacking: here are the real risks and how to harden your server.

AI 10 min read

How Much VRAM to Run an LLM Locally? The Calculation Guide (2026)

Model weights, KV cache, GGUF Q4_K_M quantization: we break down the real formula for calculating VRAM for a local LLM. Tables by size (7B, 32B, 70B), the long-context KV cache trap, and the concrete GPU mapping. No marketing approximations.

AI 9 min read

AMD Strix Halo (Ryzen AI Max+ 395) for Local AI: the €1,500 DGX Spark killer?

128 GB of unified memory, an RDNA 3.5 iGPU and a price three times lower than the DGX Spark. We break down the real LLM benchmarks of the Ryzen AI Max+ 395: actual tokens/s in dense and MoE, the 256 GB/s wall, and who this APU is really worth it for.

Security 13 min read

VPNs Under Fire: Check Point (Qilin) and Cisco SD-WAN, Two Zero-Days Exploited in June 2026

A double blow to remote-access gateways: CVE-2026-50751 (CVSS 9.3) bypasses authentication on Check Point VPNs running IKEv1 and is already being used by the Qilin ransomware, while CVE-2026-20245 opens root access on Cisco Catalyst SD-WAN Manager. Analysis, exploitation chains and remediation.

Security 12 min read

June 2026 Patch Tuesday: the largest in history, featuring a wormable kernel flaw

With roughly 208 CVEs fixed, the June 9, 2026 Patch Tuesday is the biggest Microsoft has ever shipped. At the heart of it: CVE-2026-45657, a wormable kernel RCE rated CVSS 9.8. A breakdown of the critical flaws, the zero-days, and how to prioritize, plus a sidebar on the year's 5th Chrome zero-day.

DevOps 11 min read

Veeam CVE-2026-44963: Any Domain Account Can Take Over Your Backups

CVE-2026-44963 (CVSS 9.4) lets any authenticated domain user execute code remotely on a Veeam Backup & Replication server joined to a domain. Since backups are the prime target of ransomware, this flaw deserves top priority. Analysis, scope and hardening.

AI 12 min read

Claude Fable 5: Anthropic Opens Its First Mythos-Class Model to the Public

On June 9, 2026, Anthropic launched Claude Fable 5, its most powerful model to date. 80.3% on SWE-Bench Pro, a 1M-token context window, built-in safeguards, and a Mythos 5 twin reserved for cyber defense. A full technical breakdown.

Linux 11 min read

Linux 7.1: a new writable NTFS, hardware spring cleaning, and the growing shadow of AI

The Linux 7.1 kernel lands in mid-June 2026 with a brand-new writable NTFS implementation, Landlock hardening, the removal of the i486, the Steam Deck OLED audio fix — and a deeper debate: AI-assisted contributions flooding the security mailing list.

AI 12 min read

Microsoft MAI: Seven In-House AI Models and the Looming Divorce from OpenAI

At Build 2026, Microsoft unveils MAI: seven AI models trained from scratch with no distillation, including MAI-Thinking-1 (reasoning) and MAI-Code-1-Flash (built into GitHub Copilot). Specs, benchmarks, caveats and the strategic stakes against OpenAI.

Security 13 min read

React2Shell (CVE-2025-55182): the CVSS 10 RCE haunting the React and Next.js ecosystem

CVE-2025-55182, nicknamed React2Shell, is a pre-authentication CVSS 10.0 RCE in React Server Components and Next.js. Root cause, exploit chain, state-backed actors, affected versions and remediation: everything a web developer needs to know.

Security 16 min read

n8n "Ni8mare" (CVE-2026-21858): a CVSS 10.0 flaw hands over 100,000 automation servers with no password

A technical breakdown of Ni8mare, the unauthenticated RCE (CVSS 10.0) that exposes ~100,000 n8n instances. Content-Type confusion, the full exploitation chain up to RCE, the Q1 2026 CVE wave, detection and hardening.

AI 14 min read

NVIDIA DGX Spark & RTX Spark: does the desktop AI supercomputer live up to its promise?

One petaFLOP and 128 GB of unified memory for 4,000 EUR on your desk. Behind the GB10 Grace Blackwell marketing, we break down the real numbers: memory bandwidth, actual LLM benchmarks, DGX Spark vs RTX Spark, and who it's actually worth it for.

DevOps 14 min read

Kubernetes 1.36 "Haru": User Namespaces Go GA, Ingress NGINX Retired and a Security Hardening Push

Kubernetes 1.36 "Haru": 70 changes focused on security and AI workloads. User Namespaces GA, In-Place Vertical Scaling, Memory QoS via cgroup v2, removal of the gitRepo plugin and the end of Ingress NGINX. What changes and what breaks for your clusters.

Security 13 min read

Apache HTTP/2 CVE-2026-23918: the double-free that threatens your web servers and Docker containers

Technical analysis of CVE-2026-23918 (CVSS 8.8): a double-free in Apache 2.4.66's mod_http2 enabling unauthenticated DoS and RCE. Root cause in h2_mplx.c, exploitation conditions, heightened risk on Debian/Docker, and remediation.

Security 13 min read

YGGtorrent Hacked: 6.6 Million Accounts Exposed in a Massive Data Breach

Technical analysis of the YGGtorrent hack: SphinxQL exploitation, lateral movement via SMB, 19 GB of exfiltrated data including 6.6 million accounts, payments and source code.

Security 19 min read

eBPF and Linux observability: Falco, Cilium and kernel monitoring in 2026

A complete guide to eBPF for Linux observability: Falco for runtime security, Cilium for networking, bpftrace for kernel monitoring.

DevOps 21 min read

Secrets Management in Production: Vault, External Secrets and 2026 Best Practices

A complete guide to managing secrets in production: HashiCorp Vault, External Secrets Operator, Sealed Secrets, SOPS, leak detection and an operational checklist.

Security 20 min read

Zero Trust Architecture: Principles and Hands-On Implementation in 2026

A complete guide to Zero Trust architecture: the 5 core principles, plus step-by-step implementation on Linux with mTLS, micro-segmentation and open-source tooling.

Security 10 min read

Chrome Zero-Day CVE-2026-2441: CSS Use-After-Free Actively Exploited

Analysis of CVE-2026-2441, a Chrome zero-day under active exploitation. Use-after-free in the Blink CSS engine: impact, fix, and mitigation measures.

DevOps 11 min read

Docker Compose v5: the Go SDK that changes everything for automation

Docker Compose v5, rewritten in Go, exposes a native SDK. Automation, CI/CD integration, migrating from v2, and new deployment patterns.

Security 12 min read

Ivanti EPMM Zero-Days CVE-2026-1281/1340: European Commission and Governments Compromised

Two critical zero-days in Ivanti EPMM exploited to compromise European government systems. Technical analysis, IOCs and response plan.

AI 11 min read

Mistral 3: the European open-source AI model family that changes the game

Mistral 3 brings together a family of Apache 2.0 open-source models: Small, Medium, Large. Benchmarks, local hosting, API and positioning against GPT-4o.

AI 11 min read

OpenCode: the open source coding agent that won 100,000 GitHub stars

A technical breakdown of OpenCode, the open source terminal-native AI agent. Installation, LSP, multi-session, and how it compares with Claude Code and Cursor.

AI 10 min read

Kimi Code: the open source Chinese coding agent shaking up Claude Code

Kimi Code by Moonshot AI drives the Kimi K2.5 model inside an Apache 2.0 open source CLI. Benchmarks, the PARL architecture, a head-to-head comparison and the geopolitical stakes.

AI 11 min read

Perplexity Comet: the agentic AI browser that wants to replace Chrome

A complete analysis of Perplexity Comet, the agentic AI browser: features, architecture, security, CometJacking vulnerabilities and implications.

Security 18 min read

CVE-2026-1731: Critical Pre-Auth RCE in BeyondTrust Remote Support and PRA

Analysis of CVE-2026-1731, a CVSS 9.9 flaw in BeyondTrust Remote Support. Pre-auth OS command injection, active exploitation, discovered by AI.

Security 13 min read

Fortinet FortiClientEMS: CVE-2026-21643, a Critical Unauthenticated SQL Injection

Analysis of CVE-2026-21643 (CVSS 9.1), a critical SQL injection in FortiClientEMS 7.4.4. Exploitation, RCE impact, the fix, and hardening measures.

AI 2 min read

llama.cpp RPC: Distributing LLM Inference, Yes, But Not Without Guardrails

The RPC backend in llama.cpp lets you spread inference across multiple hosts, but it is still shipped as a fragile, insecure proof-of-concept on open networks.

AI 2 min read

MLX Distributed + JACCL: A Field Guide to Mac Clusters in 2026

MLX ships distributed primitives and a low-latency JACCL backend built for Thunderbolt 5. Here's how to architect a Mac cluster the right way.

AI 2 min read

Ollama in 2026: Wiring Codex, Claude Code and OpenClaw to Local Models

Ollama's official README now highlights direct integrations with Codex, Claude Code and OpenClaw. A breakdown for a productivity-focused local stack.

AI 2 min read

Mac Studio M4 Max vs M3 Ultra for Local AI: Which One Should You Pick in 2026?

A local-LLM-focused comparison of the Mac Studio M4 Max and M3 Ultra based on Apple's official specs: unified memory, bandwidth and clustering capacity.

AI 3 min read

New LLMs to Run Locally in 2026: A Practical Shortlist for Developers

Qwen3, DeepSeek-R1, Gemma 3, Llama 3.3, DeepSeek-V3: a concrete guide to picking the right local models in 2026 and avoiding sizing mistakes.

AI 3 min read

A Mac Studio Cluster for Local AI: 2026 Architecture with exo, MLX and llama.cpp

A hands-on guide to building a Mac Studio cluster and running larger local models: sizing, Thunderbolt topology, software, limits and security.

AI 4 min read

GPT-5.3-codex: OpenAI Targets Long Coding Tasks With a More Reliable Agent

On February 5, 2026, OpenAI announced GPT-5.3-codex in Codex and the API. This model targets medium- and long-running software engineering tasks, with improved agentic behavior.

AI 3 min read

ChatGPT Tests Ads on the Free and Go Plans: What to Watch For

Since February 9, 2026, OpenAI has been testing ads inside ChatGPT on the Free and Go plans, with regional limits and contextual targeting.

AI 3 min read

Google Speeds Up Gemini in 2026: Deep Search, Gmail and AI-Assisted Shopping

In its January 2026 recap, Google announces the expansion of Deep Search, new Gemini-driven Gmail actions and an AI-enriched shopping mode.

AI 3 min read

Anthropic Raises $3B at a $183B Valuation: Maximum Pressure on the Model Race

According to Reuters (February 12, 2026), Anthropic raised $3 billion, led by Lightspeed, at a valuation of roughly $183 billion. A breakdown of what it means for the AI market.

AI 15 min read

MiniMax M2.5: the Chinese AI model that rivals Claude and GPT-5

A full analysis of MiniMax M2.5, the Chinese open-weight AI model. MoE architecture, SWE-Bench benchmarks, pricing 20x cheaper than Claude Opus 4.6, and use cases for agents and office productivity.

Linux 28 min read

COSMIC Desktop 1.0.6: the Rust-based Linux desktop gears up for Vulkan and HDR

System76 ships COSMIC Desktop 1.0.6 with file manager fixes, a native clipboard and configurable terminal hotkeys. An ambitious roadmap: Vulkan renderer, HDR and gaming.

Linux 21 min read

CachyOS: the ultra-optimized Arch Linux that's preparing a server edition

CachyOS pushes Arch Linux performance to its limits with its BORE scheduler, x86-64-v3/v4, PGO and LTO. Discover the distribution that outperforms Ubuntu by 11.6% and is preparing a hardened server edition for 2026.

DevOps 19 min read

Kubernetes 1.35: Dynamic Resource Allocation goes beta for GPUs

Kubernetes 1.35 moves DRA to beta for GPUs, marking K8s' pivot toward AI factories. Configuration, migration and Prometheus monitoring.

Linux 19 min read

Linux 7.0 Confirmed by Linus Torvalds: New Features and Release Date

Linus Torvalds announces Linux 7.0 for mid-April 2026. Live Update Orchestrator, PCIe encryption, 4x network gains: a complete breakdown of what's new.

DevOps 17 min read

Docker Kanvas: From Compose to Kubernetes Without Writing YAML

Docker Kanvas automatically converts your Compose files into Kubernetes manifests. Built on Meshery (CNCF), it takes on Helm and Kustomize with a visual approach.

Security 19 min read

Parrot OS 7.1: the pentest distribution moves to kernel 6.17

Parrot OS 7.1 arrives with the Linux 6.17 kernel, critical GRUB fixes, and a massive update of its security toolset (Metasploit, Burp Suite, Airgeddon).

AI 22 min read

Claude Opus 4.6: Anthropic Launches Agent Teams

Anthropic unveils Claude Opus 4.6 with Agent Teams, a 1M-token context window and game-changing code review capabilities. Here's what changes.

AI 15 min read

Seedance 2.0: ByteDance Launches a Cinematic AI Video Generator

Technical analysis of Seedance 2.0, ByteDance's AI video generation model. Diffusion Transformer architecture, comparison with Sora 2 and Veo 3, use cases and implications for creators.

Security 17 min read

AgreeToSteal: a malicious Outlook add-in steals 4,000 credentials via supply chain attack

Technical analysis of AgreeToSteal, the first malicious Outlook add-in found in the Microsoft Store. Vercel subdomain takeover, phishing via the Telegram Bot API and the theft of 4,000 credentials.

Security 12 min read

Apple zero-day CVE-2026-20700: critical dyld flaw patched in an emergency release

Apple patches CVE-2026-20700, a zero-day in dyld under active exploitation. A technical breakdown of the flaw, its attack vectors and the protective measures admins should take.

AI 12 min read

GRP-Obliteration: a single prompt breaks the guardrails of 15 AI models

Discover GRP-Obliteration, the technique that hijacks GRPO to misalign LLMs. A 93% success rate, with major implications for the security of AI deployments.

Security 17 min read

Critical n8n flaw CVE-2026-25049: remote code execution via sandbox escape

Analysis of CVE-2026-25049, a critical CVSS 9.4 flaw in n8n enabling RCE through a sandbox escape. Affected versions, technical exploit and hardening.

Security 11 min read

ClawHub: 341 Malicious Skills Discovered on the OpenClaw Marketplace

341 malicious plugins identified on ClawHub, OpenClaw's marketplace. A closer look at a new breed of supply chain attack targeting AI agents.

Security 12 min read

Patch Tuesday February 2026: 6 Microsoft Zero-Days and Critical Flaws to Patch Urgently

6 actively exploited Microsoft zero-days, plus Apple, Cisco and Ivanti flaws: a rundown of the critical February 2026 fixes to apply immediately.

AI 12 min read

OpenClaw: the open source AI agent that's revolutionizing automation

From Clawdbot to OpenClaw: a look back at the open source AI agent that racked up 145,000 GitHub stars in a matter of weeks. Architecture, how it works, and what it means.

Security 12 min read

Securing SSH: 8 Concrete Measures Beyond the Password

A practical guide to hardening the SSH configuration of your Linux servers. ED25519 keys, 2FA, fail2ban, port knocking, auditing and advanced logging.

Security 11 min read

DDoS Record: 31.4 Tbps, the AISURU Botnet Shatters Every Record

The AISURU/Kimwolf botnet launched a 31.4 Tbps DDoS attack, an all-time record. Analysis of the attack and protection measures for your infrastructure.

AI 11 min read

OpenAI Is Retiring GPT-4o: What It Means for Developers

OpenAI is sunsetting GPT-4o on February 13, 2026 in favor of GPT-5.2. Impact on APIs, migration strategies and alternatives for developers.

AI 13 min read

NanoClaw: the lightweight, secure alternative to OpenClaw

NanoClaw boils an AI agent down to 500 lines of TypeScript with container-based isolation. A technical look at this security-first alternative.

DevOps 14 min read

Linux Monitoring: The Essential Metrics to Watch in Production

A complete guide to the critical Linux metrics to monitor in production: CPU, memory, disk, network, processes and alerting. Commands, thresholds and best practices for sysadmins.

Security 14 min read

AI in the SOC: 30% of Cybersecurity Workflows Automated by the End of 2026

AI agents are transforming security operations centers. 30% of SOC workflows will be automated by the end of 2026. Impact, tools and risks for security teams.

AI 11 min read

Claude Code: the AI that codes inside your terminal

Anthropic's Claude Code reshapes software development with an agentic AI assistant living right inside your terminal. An overview of its features and use cases.

Linux 7 min read

5 Linux Commands Every Sysadmin Should Know

From ss to journalctl by way of lsof, these commands will save you precious time in day-to-day system administration.

AI 12 min read

Vibe Coding: When AI Writes the Code for You

Vibe coding is reshaping software development in 2026. How AI tools like Claude Code and Copilot are changing the way we write code, and what it means for developers.

AI 12 min read

MoltBook: the social network where only AIs are allowed to post

MoltBook is the first social network exclusively reserved for AI agents. 2.5 million bots registered, humans can only watch. An analysis of an unprecedented phenomenon.

Linux 12 min read

Linux Backup Strategies: The 3-2-1 Rule in Practice

A complete guide to setting up a robust backup strategy on a Linux server with rsync, rsnapshot, borgbackup and cron automation.

AI 15 min read

Autonomous AI agents in 2026: a snapshot of an ecosystem in turmoil

From OpenClaw to CrewAI, autonomous AI agents are booming in 2026. Frameworks, open source projects, security challenges: a complete overview of the ecosystem.

Security 9 min read

AI Agent Security: The Risks Every Admin Needs to Know

Vulnerabilities in OpenClaw, an exposed MoltBook database, prompt injections: AI agents introduce a whole new class of risks. A practical guide to securing them.

Security 10 min read

Why Fail2ban Is Not Enough to Secure Your Server

Fail2ban is a good start, but it only protects against brute force. Discover the complementary security layers that are essential for a production server.

AI 13 min read

DeepSeek and Open Source LLMs: Hosting Your Own AI in 2026

DeepSeek, Llama, Mistral: open source LLMs now rival proprietary models. A guide to understanding the ecosystem and hosting your own AI.

DevOps 12 min read

Nginx in Production: 7 Optimizations That Change Everything

A complete guide to optimizing Nginx in production: workers, compression, caching, buffers, HTTP/2, rate limiting and monitoring. Battle-tested, ready-to-deploy configurations.

DevOps 7 min read

Docker in Production: The Mistakes I See Most Often

Root containers, unpatched images, badly mounted volumes... A rundown of the classic mistakes and how to avoid them in your Docker deployments.

DevOps 12 min read

Ansible for sysadmins: automate without breaking everything

A practical Ansible guide for system administrators: inventory, playbooks, essential modules, roles and best practices to automate your Linux infrastructure with confidence.

Security 9 min read

Security checklist: 10 things to verify on any Linux server

An actionable checklist of 10 essential security checks to run on any production Linux server.