If you've been hearing about Claude Cowork but weren't sure what the hype is about—I just published a deep-dive that breaks it down. This isn't another surface-level "AI will change everything" take. It's a builder's guide to understanding what Cowork actually does, how the plugin system works, and what makes it different from every other AI tool claiming to automate your workflow. What's inside: - Explanation of how Cowork works under the hood (task analysis, execution, sub-agents) - The plugin architecture—11 official plugins covering Sales, Legal, Finance, Marketing, Data, Product Management - Real use cases beyond the demo videos (file ops, report generation, workflow automation) - How to build custom plugins for your specific workflows - What to watch out for (token usage, security, limitations) - Why this matters for anyone building with AI Written for AI enthusiasts and automation builders who want to understand the tech, not just the marketing. If you're experimenting with agentic AI, considering Cowork for your team, or just curious how Anthropic is approaching autonomous task execution—this is for you. Drop your Cowork experiences or questions in the comments. Especially interested in hearing from folks who've built custom plugins or integrated it into existing workflows.

A friend recently asked me how I’ve gone about naming my businesses — and whether they were overthinking the whole process. The more I thought about it, the more I realized something -- I haven’t used one strategy. I’ve used different types of naming for different purposes. One name was emotional and story-driven. One was intentionally clear and strategic. Each sends a different signal. So I’m breaking this into a 5-part series for AI Agencies to consider when naming a company: 1️⃣ Meaning-Driven Names 2️⃣ Clarity-Driven Names 3️⃣ SEO-Driven Names 4️⃣ Future-Proof Names 5️⃣ Taglines (The Hidden Multiplier) If you’re building an AI agency, your name is signaling something — And in my opinion, because that name will likely be with you for a long time, it might actually be worth a little “overthinking.” Not paralysis. Just intentional thought. Series starts tomorrow. Thanks to @Teddy Bui, @Nick Mohler and @Usman Mohammed for inspiring this series.

I’ve been using ChatGPT consistently for prompting and idea generation. I’ve also started learning n8n to build smarter automations. But if I’m honest, I don’t feel like I’ve fully mastered either. Part of me feels like I need to go back to basics, to actually master the fundamentals. For those of you who feel truly confident with automation, what did you focus on first?

I spent last week watching people struggle with AI-generated marketing content. The problem wasn't the AI. It was the prompts. "Write a blog post about X" produces 800 words of nothing. "Write a 1200-word blog post for skeptical solo consultants who think AI is too complicated, with 3 ROI examples and a FAQ addressing 'I don't have time to learn this'" produces something you can actually publish. So I built a training resource that teaches the framework I use for every marketing prompt: Context → Who is this for and what's their current state? Constraint → What's NOT allowed? Output Format → What should the structure look like? Success Criteria → How do we know it worked? It's an interactive HTML guide with good/bad examples, copy-paste templates, and a pre-flight checklist. No app to install, just download and open in your browser. If you're tired of editing below average output for 30 minutes, this will save you time. Link below. Let me know if it helps.

Give a like if you relate!