And save context." Let me be real with you: Haiku is terrible at that. Every time you try to run a task, a skill, or anything that requires actually doing something, Haiku thinks it's executing it, then in the background it just... tells you to wait. It doesn't actually trigger tools. It might be fine for writing code snippets, looking up information, or things that don't require activating any real tooling inside your ecosystem or connected apps. But using it for execution? Headache guaranteed. Wasted time. Burned tokens trying to make it understand. For that, use Sonnet. Keep Haiku for drafting something you want Sonnet to review later, that way you're not burning through your context. Or for verifying information. But don't use it to trigger tools. It doesn't do it well, and the more complex the skill, the worse it gets.

Video editing just changed forever. What used to take motion graphics artists and editors hours of manual work can now be done in minutes with natural language, no code required. In this video, I'm breaking down two methods: Claude Design for spinning up custom motion graphics through conversation, and Claude Code connected with Hyperframes for a more advanced workflow with serious customization, so every output matches your brand's tone and style. I'll walk through real examples, use cases, and exactly how to set everything up. Plus, I'm giving away the free skills and GitHub repo I use in this video so you can skip the setup and start creating right away. GITHUB REPO

If you're hitting session limits in Claude Code, this video breaks down exactly how tokens actually work and the habits that will stop you from burning through them. I cover context rot, manual compaction, the rewind feature, sub agents, markdown conversions, and a free token dashboard I built so you can see where your tokens are really going. By the end you'll know when to clear, when to chain sessions, and why the 1 million token window is insurance, not a goal to fill. Token Dashboard 10 GitHub Repos: https://x.com/DeRonin_/status/2045420155434320270?s=20

In this video, I walk you through exactly how I built my own AI executive assistant using Claude Code, and how you can do the same. We go through four phases: setting up the project, adding context and rules, building out your first skills and sub-agents, and how to let it grow over time by layering in more skills, memory, and context as your needs evolve. By the end, you'll have a clear blueprint for setting up a personal AI assistant that actually works the way you work, not just a generic chatbot, but something that knows your business, follows your systems, and gets smarter the more you use it.

Honestly, I've fallen asleep with this trend. Question, have any of you built conversational chatbots for companies using n8n? I've been making them for a while, but just for practice and demos. But at a production level, I haven't done any until now that I have the opportunity to test one. Still, I'm not sure if n8n is ultimately the best option for it. This chatbot needs to answer questions about the company and its programs, and schedule appointments. And I already have a large part of it built. But relying on a single prompt to do all that doesn't seem effective to me. But if I split it into several agents, then it's too many calls, which makes it not very cost-effective. So I'm in that dilemma.