I have been a Claude Code user for a while now, but thought the 7 day challenge would be a useful exercise in more firmly embedding the ideas. I built the newsletter, and my first topic was about the OpenAI Proposed "Robot Tax" - One thing I learned: Since I am a daily user, I actually just realized how much better Claude Code has gotten. Understanding all of the credentials helps a ton with speed, but what I built today would have taken a weekend 6 months ago. - One thing you'd improve: I woud spend more time up front really defining the outcomes I want. The end result was really good, but I could have built this using brand guidelines, voice, and potentially a component that split the output into reuable assets on different platforms. I also used this to test the new GPT Image model API, and the results are quite good!

I am working with a self hosted n8n Enterprise client. They are a Microsoft shop and looking to allow their users to create automations with their outlook accounts. Does anyone have experience with setting up vars to handle company wide secrets that allow the end user to just oauth in without touching secrets? Thanks!

Good morning, afternoon or evening. What’s amazing? I’m Jim and I’m super glad to meet you all. I am a builder and a seller. I was honored to have just become a certified n8n Expert Partner, and I love vibe coding. My favorite thing is to combine the two to deliver world class solutions to SMBs and most recently Enterprise clients. I would love to connect, make some new friends and serve however I can. Cheers!

In this video I walk you through how to use Gemini's new File Search API to take your n8n RAG agents to the next level. This setup is super simple and does not require any coding. You can drop in a document, let Gemini handle all of the indexing and storage for a very cheap price, and start running powerful RAG agents right away. The best part is that you can store your documents for free and then plug your AI agents into the File Search API so they can search through everything. It makes building RAG systems easier than ever before. I also cover a few important things to think about before you start. It is a game changer, but it is not magic. There are a few details that matter for accuracy and performance, and I walk you through those so you know how to get reliable results. COPY THE GOOGLE SHEET HERE

Two new powerhouse models from Anthropic: Claude Opus 4 - World's best coding model (72.5% on SWE-bench), excels at complex multi-hour tasks and agent workflows. Built for sustained performance on demanding projects. Claude Sonnet 4 - Major upgrade from 3.7 with superior coding (72.7% SWE-bench) and reasoning. Perfect balance of capability and efficiency for everyday use. Both models offer hybrid modes: instant responses OR deep thinking. Same pricing as before. Check out the post here