Skip to content

Search

ESC

// Blog

All Posts

34 posts

Browse all posts below or browse by topic.

10 min

Perplexity open-sources SPACE, its microVM sandbox for agents

Perplexity open-sourced SPACE, the Firecracker sandbox behind Computer. The design is solid, but its red-team report on sandbox egress is the part to read.

Read
ai-agents security infrastructure
6 min

Omarchy Quattro: the Linux distro that agents build and test

Omarchy 4 makes the coding agent a system default, is built and tested by agents, and is now funded partly in model tokens. What changed and why it matters.

Read
ai-agents ai-coding-tools dev-workflow
9 min

AgentID: Sign in with Google, Except the User Is an AI Agent

AgentMail's AgentID lets AI agents sign in to apps as their own inbox over plain OpenID Connect, and tells registered apps which human owns each agent.

Read
ai-agents oauth security
8 min

Cerebras CS-4 Is Fast. Its 30x Uses a Mid-Pack GPU Baseline

Cerebras's CS-4 puts three WSE-3 wafers, clocked twice as fast, in one rack. I measured its 30x claim: the GPU behind it does about 180 tokens a second.

Read
hardware infrastructure benchmarks
12 min

Pi Durable: An Agent Harness That Survives Its Own Crashes

Pi Durable is Earendil's experimental TypeScript harness for agents that survive crashes, fork, compact in the background and take new code while running.

Read
ai-agents llm-integration javascript architecture
10 min

Claude Code Mods: Five Tips From the Mods People Built

Claude Code mods are TypeScript hooks that redraw the UI, hold tool calls and add panes. Five tips, with real mods, for writing and installing them.

Read
ai-coding-tools claude-code anthropic dev-workflow
10 min

Cloudflare's cf CLI: The Whole API, and a Clock on Wrangler

Cloudflare's new cf CLI covers about 3,000 API operations and is built for agents. I ran cf migrate on this blog; here's what it means for Wrangler.

Read
cloudflare dev-workflow ai-agents infrastructure
10 min

WebMCP Hands-On: A Search Tool for Browser Agents

I gave this blog a WebMCP search tool to see what Chrome's agent API takes in practice: the code, the Chrome 153 quirks, who calls it, and why WebKit says no.

Read
ai-agents mcp javascript security
12 min

Orca: The Open-Source ADE for Running Coding Agents in Parallel

Orca runs Claude Code, Codex and 40 other CLI agents in parallel git worktrees, free and MIT-licensed. What it does, how it compares, and where it's rough.

Read
ai-coding-tools ai-agents dev-workflow multi-agent
6 min

Grok 4.7, Opus 5.5, GPT-6 Sol and Luna: Four Models in Two Days

xAI, Anthropic and OpenAI shipped Grok 4.7, Claude Opus 5.5 and GPT-6 Sol and Luna in two days. Prices, capabilities, caveats and how to choose.

Read
llm-integration ai-coding-tools benchmarks anthropic openai xai
10 min

Herdr: A Practical Guide to Running Coding Agents

Get started with Herdr: organize coding agents, isolate work with Git worktrees, automate reviews, and keep sessions running over SSH.

Read
ai-coding-tools ai-agents dev-workflow
22 min

GPT-6 Astra vs ARC-AGI-3: One Model, Two Very Different Scores

ARC Prize says OpenAI's GPT-6 Astra scores 62.7% on ARC-AGI-3 with its standard harness and 99.9% with OpenAI's own memory settings. What the gap means.

Read
ai-fundamentals machine-learning ai-agents benchmarks openai
19 min

LZ Caught One Event It Can't Explain. Is It Dark Matter?

A mile underground, LUX-ZEPLIN logged a single 248 keV recoil that no background it has studied can explain. What the 2.6-sigma hint means, and what it doesn't.

Read
physics dark-matter science statistics
29 min

Claude Fable 5.1: One Model, Two Names, Three Breaking Changes

Anthropic's Fable 5.1 ships with 75% cheaper cache reads, a safeguards layer your code must handle, and API changes that break history-editing harnesses.

Read
anthropic ai-coding-tools claude-code llm-integration ai-agents benchmarks
8 min

Dyson CameraJet: A £420 Toothbrush With Computer Vision

Dyson's first toothbrush pairs a 100,000-pixel camera with machine learning to spot gaps and jet mouthwash at them. How it works, and how rivals compare.

Read
computer-vision machine-learning hardware consumer-tech
6 min

MCP Grows Up: What's Changing in the 2026-07-28 Spec

The biggest MCP revision since launch goes stateless, formalizes extensions, hardens OAuth, and deprecates three core features. What matters in production.

Read
mcp protocol ai-agents infrastructure oauth
5 min

Zig in 2026: Colorless Async I/O and the Road to 1.0

Zig 0.16 lands colorless async I/O, 30ms incremental rebuilds, and a maturing toolchain. What changed, what broke, and how close 1.0 really is.

Read
zig systems-programming programming-languages
21 min

Inside Claude Code's Dynamic Workflows

Claude Code can now write its own multi-agent harness — fanning out subagents, verifying their work, and returning one answer. A sourced deep dive.

Read
ai-coding-tools claude-code ai-agents multi-agent dev-workflow
20 min

HTTP/2 Bomb: Remote Memory-Exhaustion DoS (CVE-2026-49975)

A single HTTP/2 connection from a home laptop can pin tens of gigabytes of server RAM in seconds. How the AI-discovered HTTP/2 Bomb works and how to defend.

Read
security http2 cve dos web-servers
23 min

CVE-2026-44578: Next.js WebSocket SSRF

High-severity (CVSS 8.6) SSRF in the Next.js WebSocket upgrade handler lets unauthenticated attackers proxy GETs to internal services on port 80.

Read
security nextjs cve ssrf websocket
9 min

TanStack npm Supply Chain Attack: Detect, Fix, and Recover

On May 11, 2026, attackers published 84 malicious versions across 42 @tanstack/* npm packages. Here is how to detect compromise and recover safely.

Read
security javascript supply-chain tanstack npm ci-cd incident-response
5 min

LiteLLM Supply Chain Attack: How to Check If You're Affected

Malicious LiteLLM versions were published to PyPI on March 24, 2026. Here's what happened, how to check if you're affected, and what to do.

Read
security python supply-chain litellm pypi
2 min

Getting Started with AI-Assisted Development

A practical guide to using AI coding tools effectively in your development workflow.

Read
ai-coding-tools productivity
3 min

The Next Generation of Developer Tools

How AI-native IDEs, agentic workflows, and autonomous coding are reshaping software development.

Read
ai-coding-tools productivity ai-agents
3 min

Engineering Responsible AI Systems

Practical techniques for building AI systems with bias detection, safety guardrails, and monitoring.

Read
ai-ethics architecture code-quality
3 min

Fine-Tuning Models for Your Domain

When and how to fine-tune language models using LoRA, including dataset preparation and evaluation.

Read
machine-learning tutorials ai-fundamentals
3 min

Designing AI Agent Architectures

Patterns for building reliable AI agents with tool use, memory, and multi-step reasoning capabilities.

Read
ai-agents architecture machine-learning
3 min

Running AI Models Locally: A Complete Guide

How to run open-source LLMs on your own hardware using Ollama, llama.cpp, and quantized models.

Read
ai-fundamentals tutorials self-hosting
3 min

AI-Powered Testing: Beyond Unit Tests

How AI transforms software testing with automated test generation, property testing, and intelligent fuzzing.

Read
testing ai-coding-tools code-quality
3 min

Building RAG Systems from Scratch

A step-by-step guide to building retrieval-augmented generation systems with vector databases.

Read
ai-fundamentals machine-learning tutorials llm-integration
2 min

Automated Code Review with AI

How to integrate AI-powered code review into your CI/CD pipeline for consistent quality checks.

Read
ai-coding-tools code-quality dev-workflow
2 min

The Developer's Guide to Prompt Engineering

Practical prompt patterns and techniques for building reliable AI-powered features in production.

Read
prompt-engineering ai-coding-tools productivity llm-integration
3 min

How Large Language Models Actually Work

A clear explanation of transformer architecture, attention mechanisms, and tokenization for developers.

Read
ai-fundamentals machine-learning