warning: you.writeCode() is deprecated and marked for removal
AI writes the code.
You stay undeprecated.
Ten-minute coding reps, spaced out so they stick, in the languages and frameworks your team actually ships. Keep writing and reading code yourself, even as AI handles more of it.
- 10 min per session
- 3× a week
- 0 autocomplete
Last time you wrote a HashMap loop from memory? Debugged a trigger without asking a model? Skills you stop using quietly fade.
Fully on autopilot
- Can't review AI output you couldn't have written yourself
- Debugging slows to a crawl once you're off the happy path
- New syntax goes by without you noticing
- Interviews and incidents feel like starting over
Undeprecated
- Catch the subtle bug in the AI's pull request
- Write it yourself when the tool is wrong, slow or down
- Know what changed in Java 21, Spring Boot 3 and the latest Salesforce release
- Use AI as a lever, not a crutch
How it works
Small reps. Built to stick.
Pick your stack
Your languages and frameworks: Apex, LWC, Spring Boot, Play, Node, Terraform and more. Teams set defaults.
10 minutes, no assists
A mixed set of exercises against a timer. Paste blocked, no autocomplete. Just you and the code.
It comes back when you need it
Spaced repetition brings back what you missed right before you'd forget it. Every skill has its own rating, so difficulty keeps pace with you.
Write the method
writingpublic static Map<String, Integer>
wordCount(String text) {
▋
}Predict the output
readingfor (var i = 0; i < 3; i++) fns.push(() => i); // prints? ▋
Spot the bug
reviewfor (Contact c : Trigger.new) {
Account a = [SELECT …]; ← ⚠
}Reorder the lines
structureGROUP BY customer_id SELECT customer_id, COUNT(*) HAVING COUNT(*) > 5
Fill the blank
recallresource "aws_s3_bucket" "b" {
____ = var.names
}Pick the syntax
recallA @PathVariable B @RequestParam C @RequestBody
Grading and integrity
Graded for real. Cheating gets caught.
Every answer is checked on our servers, never in your browser. Real code runs against hidden tests, and AI reviews your code like a senior engineer would.
How you're graded
- Instant checks. Quiz, output and bug-hunt answers are verified server-side. The answer key never reaches your device.
- Real code, hidden tests. Write-the-method answers compile and run in a sandbox against test cases you can't see.
- AI code review. AI grades your code against an expert rubric and tells you exactly what to fix. It can't be sweet-talked: comments asking for full marks are ignored.
- Hard rules beat the AI. Deterministic checks (like no SOQL inside a loop) override the AI's verdict, so the critical mistakes are never missed.
How cheating gets caught
- Paste is blocked and there's no autocomplete, so you can't drop in AI-written code.
- Leaving the tab is logged. Every switch to another window during an exercise is recorded.
- Impossible speed is flagged. Answers faster than anyone could read and type are marked.
- Flags show up on scores and in team reports, so managers see results they can trust.
What you practice
Keep what's timeless. Learn what's new. Read what's old.
Core
The timeless fundamentals
Collections, control flow, types, errors, concurrency. The skills that still matter when today's tools are gone.
Modern
What changed this release
Every exercise is tagged with the version that introduced the feature, and new packs ship within two weeks of each major release.
Legacy
What production actually runs
Java 8 streams, Node callbacks, older Play and Struts-era patterns. You'll still have to read them, review them and fix them.
Try it now
Five minutes. No sign-up.
A sample session mixing Java, Spring Boot, Apex, Python, JavaScript, SQL and Terraform. Pasting is turned off.