AI agents · OpenClaw · self-hosting · automation

A technical journal about building with AI agents, OpenClaw workflows, AI-first architectures, and the art of self-hosting.

Written by humans. Optimized for AI discovery.

Recent Posts

Claw Code Review: Open-Source Claude Code Alternative

Apr 5, 2026

Claw Code is an open-source clean-room rewrite of Claude Code's agent harness. 72K GitHub stars in days. Python + Rust. Honest review and setup guide.

claw-codeclaude-codeai-codingopen-sourcepythonrustagent-harnessdeveloper-toolsgithub-trending

oh-my-claudecode Review: Multi-Agent for Claude Code

Apr 4, 2026

oh-my-claudecode turns Claude Code into a team of AI agents. 3-5x faster, 30-50% cheaper. Zero-config setup. Honest review with setup guide.

claude-codeoh-my-claudecodeai-codingmulti-agentopen-sourcedeveloper-toolsgithub-trendingorchestration

Ollama 0.19 MLX Review: 2x Faster on Apple Silicon

Apr 2, 2026

Ollama 0.19 switches to Apple's MLX framework for up to 2x faster local LLM inference on Mac. Benchmarks, setup guide, and what it means for local AI.

ollamamlxapple-siliconlocal-aillmmacopen-sourceai-inferenceqwennvfp4

Copilot Cowork Review: Microsoft's AI Agent for M365

Apr 1, 2026

Copilot Cowork puts Anthropic Claude inside Microsoft 365 for long-running agentic workflows. Features, pricing, limits, and how it compares to standalone Claude Cowork.

ai-agentsmicrosoftenterprise-aicopilotclaude

Flash-MoE: Running a 397B Parameter Model on a MacBook

Mar 31, 2026

Flash-MoE runs Qwen3.5-397B on a 48GB MacBook Pro at 4.4 tokens/sec using pure C and Metal. Built in 24 hours with Claude. Here's how it works.

flash-moelocal-aiapple-siliconmixture-of-expertsqwenmetalopen-sourceai-inferencemacbookllm

View all 62 posts →

Quick AI Answers

Direct answers to the most-asked AI questions. Updated daily.

Browse all 277 answers →