kodwai logo

kodwai

Get scored on how well you drive your AI coding agent

Real coding challenges you solve on your own machine with Claude Code, Cursor or Codex. kodwai scores how well you direct the agent: Direction, Outcome, Lift.

Built by Ege Hakan Karaagac · Artificial Intelligence · Developer Tools · Education

Visit website

In category: #609 of 2048 · Artificial Intelligence

kodwai is a Community listing on Launch Llama with 36 total upvotes, founded by Ege Hakan Karaagac who joined in Oct 2026, compared against 6 alternatives.

Community listing

Listing status

36upvotes

Community score

Ege Hakan Karaagac (Joined Oct 2026)

Founder

Launched on Launch Llama Oct 5, 2026 · #11 of 165 that day

kodwai demo video

kodwai

kodwai screenshot 1kodwai screenshot 2kodwai screenshot 3kodwai screenshot 4kodwai screenshot 5kodwai screenshot 6

About

Anyone can one-shot green tests now. The interesting part is how well you drive the agent.

kodwai gives you real, ticket-sized engineering challenges (a multi-tenant rate limiter, a CRDT text buffer, an idempotent ETL pipeline, a Raft-lite log) that you solve on your own machine with your own agent: Claude Code, Cursor or Codex. No browser sandbox.

How it works:

  • Pick a challenge at kodwai.com/challenges
  • Run npx @kodwai/cli@latest challenge <slug> and choose your agent
  • Solve it the way you actually work
  • Run npx @kodwai/cli@latest submit and get your score

The score is 0 to 100 across three axes. Direction: how you steer, verify and decompose. Outcome: what shipped, replayed and stress-tested. Lift: how far you beat a solo AI, not just that you passed. Every signal cites evidence from your transcript, commits and test runs, and each challenge hides a few traps a single prompt usually misses.

Free for developers: 3 scored submissions on us, then bring your own Anthropic key. Public profile, leaderboard and a weekly league that resets every Monday. Teams can also run private real-world interviews and see how candidates work with AI agents.

Competitors: LeetCode, hackerrank, CodeSignal

Follow kodwai: X (Twitter)

Use Cases

  • Practicing real coding challenges like a rate limiter or REST API on your own machine with Claude Code, Cursor or Codex
  • Measuring how well you direct an AI coding agent through planning, testing and pushing back on its output

From the founder

About me

“I kept watching good engineers get judged by tests that either banned AI or only checked whether the tests went green. Neither tells you much anymore, because an agent can make almost anything pass. What actually separates people is how they steer: the spec they write, whether they check the agent's work, how they break a problem down, and how fast they notice when it goes sideways. So we built kodwai to score exactly that, on real tickets, in the tools people already use every day.”
— Ege Hakan Karaagac, founder of kodwai

Who it is for

  • Developers who use AI coding agents such as Claude Code, Cursor or Codex and want to improve how they work with them.
  • Engineers preparing for agentic coding workflows who want scored, hands-on practice challenges.

Pricing

Free for developers.

View kodwai pricing →

FAQ

What is agentic coding, and how does kodwai score it?

Agentic coding (also called AI-native coding, or vibe coding) is building software by directing an AI agent instead of typing every line. kodwai is a platform where developers solve real coding challenges on their own machine with their own AI coding agent (Claude Code, Cursor, or Codex) and get scored on how well they direct the agent, across three axes: Direction, Outcome, and Lift.

How can I practice agentic coding or vibe coding?

Pick a real challenge and solve it on your own machine with Claude Code, Cursor or Codex, the way you normally work. When you submit, kodwai scores how well you directed the agent and shows the moments that helped or hurt. The easy one to start with is bookshelf-rest-api: npx @kodwai/cli@latest challenge bookshelf-rest-api

Is kodwai a LeetCode alternative?

For a different skill. LeetCode is the standard for classic algorithm interview prep. kodwai does not test memorized algorithms. It gives you ticket-sized project work and scores how you direct an AI agent through it. Use LeetCode for an algorithm round, and kodwai to practice the AI-enabled rounds some companies now run.

Which agents can I use?

Claude Code, Cursor or Codex. The CLI asks which one. You solve on your own machine, and the CLI only collects from the challenge folder.

Is it free?

Yes for developers. Your first 3 submissions are scored on our key. After that you add your own Anthropic API key in Settings for unlimited runs. It is encrypted and only used to score your work.

Can a single prompt that passes the tests score well?

Not really. Direction weighs the most, and a session with no spec, no checks and no redirects scores low there however green the tests are. The hidden traps catch the rest.

How do the weekly leagues work?

A new league starts every Monday at 00:00 UTC. Your first scored run puts you in a group of up to 30. Points are your best score on each challenge, times 1, 1.5 or 2 for easy, medium or hard. The top 5 move up a division and the bottom 5 move down.

Alternatives

CodensFor non-engineers: Codens turns Notion requests into tested pull requests using five AI agents sharing one org-wide credit pool.KopaiFor experts: Kopai turns your knowledge into a sellable AI agent earning per message so you scale expertise without hourly limits and keep 70% revenue.ranwhatOpen-source AI agent security: an audit trail of what Claude Code, Codex and eighteen more agents ran, the secrets they leaked to disk, and what their tokens allow.KitstarterFor developers: Kitstarter teaches Claude Code to ask questions first and ship lean, so you deploy production-quality AI agents without over-engineering.GaugeFor developer tools: Gauge optimizes your product for AI agents so they choose your tool autonomously without sales demos.Agent IslandFor developers: Agent Island monitors Claude, Codex, Grok, and Cursor sessions locally so you track AI coding progress without telemetry.

More in Artificial Intelligence

All Artificial Intelligence tools

New on Launch Llama

Launched this week · This month's launches · All products A–Z

Tags

#coding challenges#ai agent skills#claude code#cursor#codex#developer assessment#engineering benchmark#local coding practice

Ask AI about kodwai

ChatGPT Claude Perplexity Grok

Supporters

29 founders

29 founders supported kodwai on Launch Llama.

Hold On AngolaBlue ThenicsMatikiri NETmoez cherifHerexCyriacus AnaetoAndreas ChristodoulouBrian Millot
+17

Reviews

4.0 · 5 reviews

Jacob Lock
Oct 11, 2026

Very nice tool to have! Pircing page too confusing

Mariano Luque
Oct 10, 2026

What's good: Lo mejor es que Kodwai cambia los clásicos algoritmos por evaluar cómo dirijo realmente a la IA en mi propia máquina usando herramientas como Claude Code o Cursor, midiendo con precisión la estrategia, las trampas y el resultado técnico. What could improve: Lo que le falta es pulir la claridad inicial de la interfaz y explicar mejor los límites o costos de las API keys propias antes de engancharse con las ligas semanales y los perfiles.

saidbouig
Oct 5, 2026

The hands-off vs hands-on run on the same challenge (44 vs 90) makes the case better than any headline. One thing tripped me: the demo scores out of 45/45/10 while the score section leads with 50/35/15; the note about custom rubrics is easy to miss.

Operations Team
Oct 5, 2026

Interesting angle: scoring how well you direct an AI coding agent, not just the final code. Free first three submissions is a nice way to try it.

Bank K.
Oct 5, 2026

First look from the site, haven't run a challenge yet: the side-by-side hands-off vs hands-on scoring demo makes the idea click instantly. Scoring how you direct the agent, not just passing tests, is a smart angle.

For bots & AI agents

Machine-readable versions of kodwai and Launch Llama. Fetch these instead of parsing this page.

This listing (.md) llms.txt llms-full.txt ai.txt Full catalog (.md) Endpoint index API docs MCP server MCP manifest Developer docs