in early 2016 i got payed 3.5 #BTC to design, animate and code the frontend for a #CSGO pool game.
If i wouldn't sell too fast because i needed cash, this would be the highest payed job i ever did, and possibly life-changing LOL 😅😁 link below
Do you have a story like that?
My Claude Code agentic system keeps crashing when running parallel stream executions, the report says something qith Atomic as a know issue in VS code. I can run CC in plain terminal with no problems. Any insights? #VScode#anthropic#claudecode
Back in the office. I dont know who sold us on this laptop life, but i did more work in 2 days un the officw than 3weeks on a laptop travelling. #claudecode
🤖 Building agentic AI systems? Hooks are your secret weapon.
Hooks are the difference between "AI assistance" and "AI you can trust in production"
Without hooks: Manual quality checks, inconsistent standards, contaminated output, 3 languages mixed in 1 sentence.
The longer the content piece, the worst it gets - context window issue.
With hooks: Auto-triggered validation, self-iteration, quality gates that BLOCK bad content before it ships
Our 5-language content system caught 35 contaminated articles automatically. No human review needed until AI fixed them.
#claudecode #agentic #Anthropic
Developing on the go with AI !
Working in web browser Claude Code while travelling, it's an amazing & exciting new reality. #Anthropic team keeps amaze me!
Thank you guys for free 1k € free tokens, let's see if i can burn them on the go. #Claude#claudecode
Ai coding tools are just tools, pair them with coding knowledge, and you output becomes 10x....
If you are a developer - stop resisting the (r)evolution, or you will be damn sorry.
In a year there will be no-one writing code by hand. #AI#claudecode#Anthropic
Claude Code: One simple step. 10x code quality.
Most devs dive straight into coding. I did too.
Then I'd spend days untangling structural problems that could've been avoided.
Here's what changed:
Claude Code's Plan mode agents.
I've always used Plan mode to start projects. It was my routine.
But the latest update? Non-negotiable.
The agents help you think through:
Edge cases before they bite you
Dependencies before they conflict
Architecture before you build the wrong thing
It's like having a senior architect review your approach—before you write a single line.
Reality check:
Better code doesn't come from coding faster. It comes from planning smarter.
Since making Plan mode my mandatory step 1:
- Fewer bugs in production
- Cleaner implementations
- Better architecture from day one
- Less refactoring (and frustration)
The hard part?
Resisting the urge to "just start coding" because it feels productive.
Use Plan mode first. Thank me later.
P.S. What's your biggest regret from skipping the planning phase?
A slash command is basically a saved prompt that accepts variables.
Think of it like a small reusable function:
/command [arg1] [arg2?] [arg3?]
It’s fast, structured, and consistent.
You define how a task should be done once — then call it whenever you need.
I use them for SEO, content generation, code reviews, even client onboarding.
Slash Commands: turn your AI into a command-line assistant.
Everyone’s talking about “prompt engineering,”
but the real power move is command engineering.
In my Claude Code setup, I don’t “ask” anymore.
I use commands like: /seo-keyword "dental implants" 2528 "Dutch"
Here’s how I built my own AI command language 👇