Weekly Vibe – Agents, Local Models, Security, and “Where Do I Even Start?”

This week’s call was a good mix of beginner questions, deep agent architecture, and some real “where is this all going?” conversations.

We had five of us on: Wes, Aty, Shawn, Chris, and Gary — and the spectrum of experience in the room actually made the discussion better.

Here’s what’s in the video:

🧭 “I’m Not a Developer. Where Do I Start?”

Gary’s question was simple and honest:

I’ve done some HTML and CSS… but with all this B-Mad, Claude Code, OpenClaw stuff — where do I even start?

He’s running VS Code on a Raspberry Pi (which is awesome, by the way), trying to understand the stack without breaking his main machine.

We talked about:

Not needing to become a “developer” in the old sense
Starting with outcome definition instead of tools
Keeping early builds simple (MVP mindset)
Avoiding the trap of over-architecting too soon

If you’ve felt overwhelmed by:

context windows
local models
agent frameworks
“greenfield vs brownfield” talk

You’ll relate to this part.

🧠 Sonnet 4.6, Codex 5.3, and the Shift in Model Power

We got into the recent updates:

Sonnet 4.6 improvements
1M context window options
Codex 5.3 becoming very test-driven
Models increasingly self-checking and structuring output

There was a really interesting comparison between Claude and Codex:

Claude tends to “get it working”
Codex tends to enforce tests and longer-term structure

That difference matters once your projects get big.

🏗 Chris: Building an OpenClaw Alternative (Local Model Focus)

Chris shared that he’s been building his own agent framework — designed to eventually run well on local LLMs.

He’s intentionally “skating where the puck is going.”

Key themes:

Preparing for local models to get strong enough
Adding guardrails around smaller models
Running into scaling problems as projects grow
The importance of test coverage before things get out of control

If you’re building something serious, this part is worth watching.

🔐 Security, Containment, and “How Do I Not Blow Up My Business?”

Shawn brought up a real concern:

How do I experiment with agents without risking my production systems?

We talked about:

Isolated machines
Restricted environments
Limiting API key exposure
Syncing memory manually
Not connecting Shopify directly without a framework

There was a good practical discussion about containment vs experimentation.

No fear-mongering — just real operational caution.

🧩 Agents vs Skills vs Workspaces

Aty shared something interesting:

Instead of just thinking in “skills,” think in:

Agents that own specific skills.

He’s building:

Separate memory stacks
Blueprint files
Tone/style files
Reusable architecture blocks

The idea is:

Keep truth-based data clean
Keep brand identity consistent
Sync memory explicitly
Rehydrate context into new workspaces cleanly

This was more architectural than tactical — but really useful if you’re thinking long-term.

📊 Real Business Application: Using Agents for Product Strategy

Shawn talked about something practical:

Pull best sellers from Shopify
Match against customer survey data
Predict next high-performing products
Build a “staff of 10 PhD-level assistants”

But with guardrails.

The big theme:Architecture first.Then automation.

🔄 Memory, “Mile Markers,” and Not Breaking Everything

There was a great metaphor from a truck driver friend of Shawn’s:

“Mile markers.”

Instead of going back 1,000 miles when something breaks — go back one mile.

That turned into a discussion about:

Structured checkpoints
Versioning
Tests
Keeping projects from collapsing under their own weight

If you’ve ever broken your entire codebase with one change, you’ll appreciate this section.

🧪 Generative UI (Tambo Demo)

Wes showed a quick example of generative UI using a framework called Tambo.

Instead of just chat → text output, it was:

Chat → structured UI components
Auto-generated forms
Dynamic interface rendering

It’s subtle, but it hints at where app interfaces are heading.

🧠 Big Picture Themes

Some threads that ran through the whole call:

Models are no longer the limiting factor.
Architecture and clarity are the new bottlenecks.
Not being a traditional developer may actually be an advantage.
The hard part isn’t generation — it’s containment, structure, and memory.

Overall, it was a great call.

It was:

builders comparing notes
someone new asking real questions
someone deep in the weeds sharing real scaling pain
practical security discussions
architecture over excitement

If you’re somewhere between “curious” and “building something serious,” this one is worth your time.

Watch it when you’ve got an hour and want to think more clearly about how you’re structuring your stack.

1:09:01

2 comments

Vibe Coders

skool.com/vibe-coders

Master Vibe Coding in our supportive developer community. Learn AI-assisted coding with fellow coders, from beginners to experts. Level up together!🚀

Leaderboard (30-day)

+11