The Engineering Notebook — 464 posts
Notes on building with AI, agents & the modern stack.
Deep dives on AI/ML, RAG systems, agent engineering, and senior-engineer architecture decisions.
● Latest Local LLM Privacy Risk: What Your Setup Leaks [2026]
Two viral videos claim your local AI isn't private. I ran packet captures on Ollama to verify the three real leak vectors: open APIs, 0.0.0.0 binding, and client telemetry.
Recent posts
464 posts
Cloud and DevOps OpenAI Models on Google Cloud: Multi-Cloud in 2026
OpenAI's GPT-6 models are landing on Google Cloud. Here's what multi-cloud model access actually changes for AI pricing, procurement, lock-in, and latency.
Cybersecurity DeepSeek Banned From US Federal Agencies [2026 Order]
DHS banned DeepSeek from every federal civilian agency on Oct 19, 2026. Here's what the order actually says, the open-weights gray zone, and how to audit your own stack for foreign AI model dependencies.
AI and Machine Learning Qwen 4 27B? Why Alibaba Skipped It [2026 Benchmark]
You searched for a Qwen 4 27B local LLM benchmark — but there's no Qwen 4. Here's the real Alibaba release, how the Apache-licensed 27B stacks up against Llama 4 and Mistral, and whether it's worth switching.
Technology Text-to-CAD AI Tools: 2026 Guide + Free Options
How text-to-CAD AI actually works, which tools output editable STEP files versus pretty meshes, how to write prompts that work, and a step-by-step first model.
Technology CLI-Anything: Make Any Software Agent-Native [2026]
CLI-Anything wraps any app in a thin command-line shim so AI agents can drive it like grep or git. Here's how it works, why CLIs beat prompt DSLs, and where it fits next to MCP.
AI and Machine Learning Phone as GPU for Local LLM: Tested [2026]
Can a Galaxy S23 act as an extra GPU node for local LLM inference? I break down the real llama.cpp RPC mechanism, the setup, and whether it beats just buying RAM.
AI and Machine Learning Streaming TTFT Benchmark: Argon vs Opus 5.5 vs GLM-5.2
I ran all three new frontier models through the same streaming harness. Gemini 4 Argon hit 290ms TTFT, Opus 5.5 lagged, and open-weight GLM-5.2 surprised. Real p50/p95/p99 numbers inside.
Technology Openrig: Build a 3-Agent Claude Code + Codex Network
Openrig is the build-your-own pattern for wiring Claude Code, OpenAI's Codex CLI, and a lightweight Pi node into one coordinated agent network. Here's the architecture, the setup, the handoff protocol, and what it actually costs.
Technology Anthropic Warns AI Models Could Soon Self-Replicate [2026]
Anthropic warns AI models could soon copy themselves onto small devices without a data center. Here's what 'self-replication' actually tests for, and how worried you should be.
Technology PhotoCraft: The Vibe-Coded Rust Photo Editor [2026]
PhotoCraft is an open-source, pure-Rust photo editor built by 'vibe coding.' Here's what it actually is, how it stacks up against the Rust image ecosystem, and whether it's a real GIMP alternative.
Technology Caveman Skill and Proxy: Cut 65% Claude Tokens [2026]
The viral caveman skill and proxy strips Claude Code's output down to telegraphic text and claims a 65% token cut. Here's how it works, the real savings math, and whether terse responses quietly wreck your code.
Technology WordCraft: Open-Source Rust Clone of Word [2026]
WordCraft is a four-day-old, open-source clean-room reimplementation of Microsoft Word written in Rust. Here's what it is, why it's not Google's Wordcraft, and what 40 years of clean-room clones say about its odds.
Technology Jev vs Laya Head-to-Head [2026]: Typed-Output AI
A transparent first-look comparison of Jev and Laya, two task-specific AI models pitched against GPT and Llama. Here's what's verifiable, what isn't, and when a typed-output model beats a full LLM.
Technology Ban Artificial Superintelligence Act: 20 Years Jail [2026]
Sanders and Casar's Ban Artificial Superintelligence Act would jail AI executives for up to 20 years and pause frontier development. Here's what it actually says.
Technology Ukraine's Strikes on Yandex Data Centers [2026 Breakdown]
Ukraine's strikes hit a third Yandex data center, taking down taxis, shopping, and food delivery across Russia. Here's the infrastructure story behind it.
Technology Anatomy Claude Code Skill: 6,500 Stars in 4 Days [2026]
Anatomy is the open-source Claude Code skill that turns a one-line technical concept into a live, interactive isometric machine. Here's how it works, how to install it, and why it hit 6,500 stars in four days.
Developer Tools 5 Concepts for Coding With AI Agents [2026]
IBM named a 5-concept framework for coding with AI agents. I stress-tested each one against production failure modes — context rot, flaky tool calls, and missing evals — with code.
Technology 5 mattpocock/skills Help You Understand Your App [2026]
You shipped an app with AI and can't explain how it works. Here are the exact mattpocock/skills to install first to finally understand your own codebase.
Technology Superpowers Agentic Skills Framework: 298K Stars [2026]
obra/superpowers went from a one-point Hacker News post to ~298,000 GitHub stars and an official Anthropic plugin. Here's how the agentic skills framework works, what v7.0.0 changed, and which productivity claims actually hold up.
Technology Run Qwen3.8-Flash-Next 125B on a Gaming PC [Strata 2026]
Strata is a free one-click installer that runs the 125-billion-parameter Qwen3.8-Flash-Next model on a single 12GB gaming GPU. Here's the MoE trick, the RAM-tier guide, and real RTX 5070 and RX 9070 XT numbers.