I’ll definitely be giving this a swing. It’s awesome that you’re attempting to solve these problems either way.
One problem that I notice is that it’s extremely tricky to track tasks and reliably remember where I left off within Claude Code workflows.
My guess is having a more robust context-view would make it easier to come back to a re-jolt the brain with what was actually happening.
I’ve done large scale migrations across a few projects now (RoR major upgrades) and can 100% see how these would have increased the output quality and perhaps reduced the total quantity of bugs.
I’m definitely going to borrow this and start implementing my next set of upgrades leveraging them!
I’d be curious to see people create shareable workflows (more than likely prompts that can be shared) that solve a common set of migrations (like AngularJS -> angular or major nodeJS upgrades OR EVEN NodeJS -> Bun transitions).
I’m feeling this problem every single day, 12 hours per day. An immaculate encapsulation of the problem space.
Now, the question is, how will we be progressing to elevate to the next space?
@addyosmani This is probably the most well explained highlight of this year’s engineering problem I’ve seen.
Well deserved follow, and it also reminds me that I need to be thinking deeper about these problems and writing more, well-thought out articles on them.
You know, grabbing people's attention is extremely fucking difficult when you're just a little speck, whispering in a world of beastly, yelling giants.
The skill of gravitating attention is EXTREMELY powerful. It's the highest possible form of leverage you can create today.
software engineering in 2026:
- your package manager is compromised
- your cloud provider blocks your account
- github itself is hacked
software is solved