/show-me is a phenomenal skill
Makes PR descriptions extremely easy to read
Basically a toolbox of "nice ways to look at code"
Nice work, @dexhorthy
https://t.co/phB1dAPwgG
I’ve been thinking about this.
AI coding agents struggle with two things humans do naturally:
- Build a mental map of the codebase.
- Learn a team’s conventions.
Maybe we’re asking the wrong question.
Should AI work like humans?
Or should we change how we build software?
SYSTEM DESIGN PLAYBOOK
Giveaway Alert!!!
• System design fundamentals.
• Condensed notes to read before system design interview.
• Must know concepts from real-world software engineering case studies.
To get it for free:
1 Retweet & Follow @systemdesignone
2 Reply "Playbook"
Then I'll DM you the details.
This Reddit post is a great example of using AI the right way.
Learning by doing with guardrails.
AI is a tool to improve productivity, code quality, and problem solving, not a shortcut to avoid thinking.
https://t.co/rR26MhY4nG
Using Hexagonal Architecture in a side project.
Goal: keep business logic independent from infrastructure so I can swap implementations and try different stacks.
Ports & adapters make this surprisingly practical.
@TAbrodi One counterargument: Convex has more lock-in due to its proprietary backend and programming model, while Supabase is built on standard PostgreSQL.
That makes it easier to migrate away from Supabase as your app scales.
I’m on week five of trying to vibe code a replacement for some dumb saas that we use and it’s so incredibly frustrating that I’m slowly realizing it’s actually a quite complex and thoughtful piece of software.
One rule I follow when working with AI:
Never let it write code before agreeing on a plan.
When AI jumps straight into coding, it often produces something that looks correct but isn’t aligned with what you actually want.
Instead, I use this workflow.
Watching the Enterprise UI workshop live on @FrontendMasters today!
Lots of great topics lined up: microfrontends, monorepos, design systems, scaling TS and more.
Great excuse to dive deeper into large-scale frontend architecture 😄
It’s funny to see engineers putting so much effort into writing detailed prompts for AI.
I just wish we had put the same effort into writing clear tickets instead of one-liners nobody else could understand.