Pi Coding Agent: The SDK Is the Real Reason to Care
I spent a week with the Pi coding agent. The CLI is good, not revolutionary — the TypeScript SDK is what makes it worth switching. Setup, extensions, and a working build.
Thomas Wiegold - AI & Web Development Blog
Insights on AI solutions, web development, and modern software engineering practices
Latest posts
Opus 5 vs GPT-5.6 vs Grok 4.5: Four Real Build Tests
I ran Claude Opus 5, GPT-5.6 Sol and Grok 4.5 through four hard builds: gallery site, spreadsheet app, image codec, repo audit. Times, results, verdict.
Qwen3.8-Max Review: I Tested Alibaba's 2.4T Model
I ran Alibaba's new 2.4T Qwen3.8-Max-Preview through 4 real coding tests. Results rival Fable 5 and Grok 4.5 — with one big catch: speed.
AI for Tradies: What Actually Works (and What's a Waste)
A hands-on guide to setting up AI for a trade or small services business. Real workflows for email, quotes, calendar and invoices — no hype.
Grok 4.5 Review: I Tested SpaceXAI's Cheap Coder
I ran Grok 4.5 through my usual coding tests — website builds, a Go poker sim, a site audit. It's fast, cheap, and it found a bug no other model caught.
MAI-Code-1-Flash Review: I Tested Microsoft's First Coding Model
I tested Microsoft's MAI-Code-1-Flash coding model on real projects. Fast and cheap, yes, but here's why I won't be switching from Kimi K2.7 Code.
Context Engineering for AI Agents: A Field Guide
AI agents drift, forget, and derail on long tasks. Learn context engineering — 8 practical rules to keep your agents reliable, grounded, and on-goal.
When Is DeepSeek V5 Coming Out? The Honest 2026 Answer
DeepSeek V5 has no announced release date. Here's what the July 24 deprecation actually means, plus V4 Pro pricing, vision API status, and Claude vs GPT-5.
Hermes Agent: What It Is and Why Devs Switched
Hermes Agent is the self-improving autonomous agent devs are switching to. Here's what it does, how it compares to OpenClaw, and where it falls short.
Claude Fable 5 Review: Best AI Coding Model Yet
My hands-on Claude Fable 5 review. I ran my usual coding tests and it one-shotted a poker sim no model ever beat. Best coding model yet, with caveats.
How to Build MCP UI: An AI Agent Approval Gate
Build custom MCP UI: a human approval gate that makes your AI agent pause for Approve, Edit or Reject before it acts. Full TypeScript + React guide.
MiniMax M3 Review: Finally Matching GPT-5.5 & Opus?
I ran my usual coding tests — two websites, a poker sim, and a code audit. Here's how MiniMax M3 actually stacks up against GPT-5.5 and Opus 4.8.
Google Antigravity 2.0 Review: I Tested Gemini 3.5 Flash
Hands-on Antigravity 2.0 review: I tested Gemini 3.5 Flash on real coding tasks. Fast, impressive design — but the hidden token cost changes everything.
Notion Workers for Small Business: A Hands-On Guide
Notion just launched Workers and a CLI — finally a real automation layer for small businesses. What shipped, what it costs, and a hands-on build.
Claude Code Hooks: From Linting to Hardened AI Workflows
Claude Code hooks make your AI agent deterministic. Hands-on guide covering formatting, security, logging, and forced verification with TypeScript.
DeepSeek V4 Review: I Tested It on Real Code
DeepSeek V4 is here. I ran it through a TypeScript codebase audit, a poker simulation, and two web designs. Here's how it really compares to Opus 4.7 and GPT-5.5.
The Ralph Loop: How Recursive AI Agents Actually Work
How the Ralph Loop turns Claude Code, Codex /goal, and any LLM into a recursive AI agent that ships code overnight — and when it actually works.
Build an AI SEO Agent in TypeScript with Claude
Code a real AI SEO agent in TypeScript — crawls competitor sites, scores pages with Claude, returns the most relevant content. Full tutorial.
How to Use the Claude Code Frontend-Design Plugin to Stop Shipping AI Slop
The Claude Code frontend-design plugin fixes generic AI design. Here's how it works, how to prompt it, and how to apply it to WordPress and Shopify themes.
OpenCode Go Review: Is the $10 AI Coding Plan Worth It?
OpenCode Go gives you 7 Chinese AI coding models for $10/month. After a week of real use, here's what works, what doesn't, and who it's actually for.