My boss gave me some advice today that I didn’t want, but it’s what I needed. Operational work (budgets, coordination, etc) saps my soul, but there’s a “right way” to present info to higher ups. Lesson learned!
[blog] Enable on-demand expertise with Agent Skills in Genkit Go. I built my first thing with Genkit Go yesterday. It’s the best AI SDK for Go and has some cool features like this.
[article] Today’s Codex will feel “primitive” by fall — and its own team’s roadmap backs it up. What’s the next leap forward for these AI builder tools? I don’t know. But it’s coming fast.
[blog] Agent Plugins package your skills, tools, and more. I’m excited about this. Now we can better group together the skills and MCPs that address a given solution set.
[article] Should You Self-Host Inference? Important question, and here’s some input for your heuristic. This seems pretty reasonable to me.
[blog] A new way to website. Static sites aren’t new, but we can throw away many old assumptions about how to create and run our ideas.
[blog] Scaling AI Agent Infrastructure with the MCP Stateless updates. Excellent post from those who led the charge to decouple the protocol from the transport. Big implications!
[blog] Introducing Muse Code and Muse Spark 1.2. Everybody gets a coding CLI! They’ve got a strong feature set and early benchmark numbers.
[article] What nobody tells you about writing agent skills. Great advice. Read this to gather some battle-tested knowledge about what to include in your skills, how to maintain them, and whether you even need one.
[blog] Agent and Model Evaluations in Gemini Enterprise Agent Platform are now GA. This looks great. I started to try it out yesterday, but ran out of time and couldn’t do it justice. I’ll put together a blog post at some point.
[blog] Simon Willison on Technical Blogging. Great advice. Just write. Most people will never know what you skipped to get it out the door.
Want to get this update sent to you every day? Subscribe to my RSS feed or subscribe via email below:
Leave a comment