Unlearn @UnlearnDev

Built for developers who want to unlearn outdated defaults and start building engineering judgment. Powered by @Bitter_Brains unlearn.dev Joined March 2017

Tweets

5K
Followers

45K
Following

165
Likes

4K

Certificates.dev @CertificatesDev

3 days ago

Not sure if you're ready for the AI-Native Developer Certification? The free prep review covers all six exam domains. Check it out, no purchase needed. 📖 vueschool.notion.site/AI-Native-Dev-…

1 1 2 364 0

View Details

Unlearn @UnlearnDev

5 days ago

In the age of agentic engineering, refactoring is disappearing from git history. GitClear's latest report puts moved lines at 3.8% of code changes so far this year, down from 21% in 2022. Agents write a lot of code, but they don't go back and tidy up unless you ask them to.

0 0 3 1K 0

View Details

Unlearn @UnlearnDev

5 days ago

@marcelpociot Looks cool!

0 0 1 177 0

View Details

Certificates.dev @CertificatesDev

5 days ago

The response since pre-orders opened has been better than we expected. 🙌 A lot of developers want to be among the first to prove they can direct AI with real engineering judgment, not just ship what it writes. Good company to be in. 2 hours left on the Early Bird. ⏰ ai.certificates.dev

0 2 4 323 0

View Details

Unlearn @UnlearnDev

a week ago

Your tests can stay green while your app breaks. Ask an agent to tidy up a file with no rules in place, and it'll happily change behaviour. We show exactly this at the start of Workflow 7 (Refactoring with an AI Agent), then spend the rest of the workflow making sure it doesn't happen to you. Start unlearning today. unlearn.dev

0 0 1 1K 0

View Details

Certificates.dev @CertificatesDev

2 weeks ago

Mark your calendars and gain free access to Vue.js Mid-level training 📣 📆 Sept 26-27 💡 Reactivity, Composition API, Vue Router + more 🛠️ 20 coding challenges 📝 Trial exam Developed with @evanyou, the creator of @vuejs. Save your free seat👇 go.certificates.dev/vfw6x

0 3 4 384 0

View Details

Unlearn @UnlearnDev

3 weeks ago

Most people decide what to refactor by opening a file and not liking how it looks. 💡 A better starting point? Files you or your agent have to keep going back into. They cost time and money. Use this to find them: git log --name-only --pretty=format: -- 'app/*' | grep -v '^$' | sort | uniq -c | sort -rn | head -20

0 0 2 1K 2

View Details

Frontend Nation @frontendnation

3 weeks ago

Announcing TypeScript Day 🔥 Featuring talks built entirely around TS, with real-world insight from industry experts who use it every day. 🗓️ Oct 20 ⏰ 3 PM CEST 📍 Online & Free Get tips and community connections to sharpen your TS skills. Sign up 👇go.frontendnation.com/ts

0 3 3 456 0

View Details

Unlearn @UnlearnDev

3 weeks ago

AI writes race conditions all the time. They’re some of the hardest bugs to spot in review because the code often looks perfectly reasonable.

0 0 1 2K 0

View Details

Certificates.dev @CertificatesDev

4 weeks ago

🔴 We are live in less than an hour with @alexdotgs for a free, live AI session with hands-on exercises and a Q&A. There is still time to join - sign up below, and the link will be sent to your inbox. ai.certificates.dev/events/design-…

0 2 4 576 0

View Details

Unlearn @UnlearnDev

4 weeks ago

Learn more about Spec-driven development with OpenSpec in our deep dive: unlearn.dev

0 0 2 231 0

View Details

Unlearn @UnlearnDev

4 weeks ago

Finally you archive it, folding the change into the living spec so the next one starts from what the app actually does.

1 0 1 210 0

View Details

Unlearn @UnlearnDev

4 weeks ago

Spec-driven development is pretty simple: write down what the change is and how it should behave before the agent writes any of it. 🧵

2 2 3 2K 0

View Details

Unlearn @UnlearnDev

a month ago

Most people write an agent skill for a topic they think is important. A better starting point is something your agent keeps getting wrong.

1 0 3 2K 0

View Details

Unlearn @UnlearnDev

a month ago

Most of the time you spend on a bug isn't spent fixing it. It's spent working out what's actually wrong in the first place. That's the part we hand over to an AI agent in Workflow 6 (Debugging and Performance). Learn more 👇 app.unlearn.dev/workflows/fixi…

0 0 0 1K 0

View Details

Unlearn @UnlearnDev

a month ago

We wrote up the full breakdown, what goes into context, why bigger isn't better, and the three techniques (just-in-time retrieval, compaction, sub-agents) that keep a long session from falling apart. Read more: app.unlearn.dev/blog/context-e…

0 0 1 210 0

View Details

Unlearn @UnlearnDev

a month ago

If you've used Claude Code, you've already watched this play out. CLAUDE.md is the system prompt and memory in one file. Grepping the repo instead of reading it all is just-in-time retrieval. "Compacting conversation" is compaction, exactly.

1 0 1 251 0

View Details

Unlearn @UnlearnDev

a month ago

Every time you send a request to your agent, it starts from zero. No memory of your last message, your project, or you. Everything it needs has to be packed into that one request. That's the context window, and deciding what goes into it is the real skill now.