Agent-First AI Plumbing

· The Fluency Briefing

The Fluency Briefing

Your Guide to What's Happening in AI and Why It Matters to You

Monday, November 9, 2026


Newsletter header image

Google shipped two new TPU chips today, one of them built to keep secrets, while two separate startups open-sourced tools that let AI agents write code and probe for security holes on their own. Add Automattic giving vibe-coded apps a real home, and you get one clear shape: the plumbing under AI is being rebuilt for agents first and humans second.

Today in AI:


Section break image

Today's Takeaway:

Reason Machines built Aspen because standard Git chokes when agents push commits nonstop, and 0sec's harness lets those same agents attack your systems to find the weak spots first. Put them together and your future software pipeline runs on two kinds of robots: one writing, one breaking in. Spacefast then hands the results a database and a login screen in minutes. Each layer speeds things up, and none of them includes a person reading the code.

That's the stake for a small business: a vibe-coded customer portal on Spacefast can go live before anyone checks how it stores passwords. Our contestable view is that the free security harness matters more than the faster Git server, because a fast pipeline only helps if something in it is checking the code. marktechpost.com | Replit


🔍 Myth Buster

The myth: ""Open-sourcing AI attack tools like 0sec's harness mostly just hands criminals a free weapon.""

The reality: 0sec Labs built '0' as a defensive harness, not an attack kit: it gives small teams the same attacker-style testing big companies pay red teams for, via a dashboard, CLI, and MCP support so AI assistants can plug straight in. The whole point is to let a business probe its own vibe-coded app, like one built on Spacefast, before it ships, closing the gap the newsletter itself flags: fast pipelines (Aspen's rapid-fire commits, Spacefast's instant hosting) only help if something checks the code.

The nuance: It's fair to worry that any automated security tool can be pointed at someone else's systems instead of your own, which is exactly why pairing it with a human asking 'where are passwords stored' still matters.


Newsletter closing image

The Bottom Line

The Pattern: Earlier this fall, agent stories were about rules and reviewers; this Monday, agents are getting purpose-built infrastructure, from Google's chips to Git servers, so the tooling is now built for machines first and reaches people later.

Our Call: At least one more vibe-coding host, beyond Spacefast, will announce built-in automated security scanning by January 31, 2027. More likely than not. We'll grade this one in a Friday digest.

Your Move: Fifteen minutes today: if anyone on your team shipped an app through Lovable, Spacefast or similar, list each one and ask its builder in writing where user passwords and customer data are stored.


What We're Working On

✨ AI Can Do That? Free 3-Day Live Workshop - Join us and Dylan Stewart live, November 27-29 at noon Pacific, as we take real ideas from attendees and work through them with AI. No experience needed, and one registration covers all three days. → Reserve My Free Spot

✨ Free 30-Minute AI Consultation - Discover how My AI Fluency can help your business unlock the potential of AI. We'll discuss your goals, explore practical AI opportunities for your industry, and outline clear next steps. → Schedule Free Call

✨ How AI-Fluent Are You? - Test your AI fluency with our interactive quiz. See how you stack up and discover what to learn next. → Take the Quiz

📞 Book a Consultation | 🌐 Website

My AI Fluency

Fluently yours, The My AI Fluency Team