Challenge catalog with difficulty filtering
Browse “real, ticket-sized problems” across multiple categories and filter them by difficulty before starting a run.
kodwai
Kodwai is a CLI-led platform for developers to solve real coding challenges with an AI agent and get scored on how they directed the session. It supports local, on-machine workflows with tools such as Claude Code, Cursor, and Codex.
What Kodwai is
Kodwai is a platform for developers to solve coding challenges with an AI coding agent and be scored on how they directed the session, not just whether the final tests passed. The product is built around real, ticket-sized problems that you run locally with tools such as Claude Code, Cursor, or Codex.
The workflow starts from the terminal: you pick a challenge, launch it with the CLI, work on your own machine, and submit a package that includes code, test runs, git history, and the agent transcript. The resulting score contributes to a public leaderboard and a visible profile.
Core capabilities
CLI-based challenge setup
Start challenges from the terminal with the Kodwai CLI, which downloads the problem statement, starter files, and tests, initializes a git repo, and starts the timer.
Local agent-driven workflow
Work on your own machine in your own editor, using your own agent, with no sandbox or artificial constraints described on the site.
Rich submission evidence
Submit a package that includes code, git history, test runs, agent transcript, and elapsed time so the platform can score the whole session.
Multi-signal scoring
Scores are broken into Direction, Outcome, and Lift, with per-signal evidence cited in the result view.
Leaderboard and profile tracking
Every scored run feeds a public leaderboard and a public profile that can be shared with others.
Practical ways to use Kodwai
Solo developer practice
Use Kodwai for timed, ticket-sized practice when you want to work through a realistic coding problem with an agent rather than a puzzle solved from memory.
Agent workflow benchmarking
Use the scoring and transcript review to benchmark how well different agent styles, prompts, or verification habits perform on the same task.
Public skill profile
Use the public leaderboard and profile to show evidence of your runs and compare progress over time.
Category-specific skill sharpening
Use the challenge catalog to practice on different categories such as backend, frontend, security, search, or mobile work.
Interview-style skill assessment
Use the platform as a structured way to evaluate how you direct, verify, and recover when an AI agent goes wrong on a coding task.
Pros and Cons
Pros
- Designed around real coding tasks rather than whiteboard-style puzzles.
- Lets you use your own machine and your own editor instead of a hosted sandbox.
- Captures more than the final answer by including prompts, recovery, test runs, and git history in submission data.
- Uses a scoring model that separates direction, outcome, and lift, which helps explain why a run scored the way it did.
- Adds a public leaderboard and profile for tracking progress over time.
Cons
- The pricing page returned 404 in the provided source, so pricing and plan structure are not confirmed on the pages reviewed.
- The available pages do not spell out every supported agent, setup requirement, or environment constraint.
FAQ
How do I start a challenge?
Kodwai is meant to be run from the command line on your own machine. The site says you start a challenge with its CLI, which downloads the problem files, initializes a git repo, and starts the timer.
Which AI agents does it support?
The homepage says you can use Claude Code, Cursor, or Codex with the platform, and also mentions “your preferred AI agent.” The rendered page does not provide a fuller supported-integrations list.
What does Kodwai score?
Kodwai is built around solving a ticket-sized coding problem, submitting your code and session data, and then receiving a score that reflects direction, outcome, and lift. The score includes evidence from the session, and completed runs can place you on the leaderboard.
Can teams use it?
The source describes a public leaderboard and a public profile that builds from scored runs. It does not show team dashboards or organization-specific features on the pages provided.
What is it best used for?
The site positions Kodwai as a way to test agent-directed coding skill on real problems rather than whiteboard puzzles or simple test-passing exercises. That makes it a better fit for practice, benchmarking, or skill demonstration than for general-purpose code execution.
Quick Facts
- Category
- Developer Tool
- Primary workflow
- CLI-led coding challenges with an AI agent
- Supported agents mentioned
- Claude Code, Cursor, Codex
- Platform type
- Web app plus local CLI workflow
- Source domain
- kodwai.com
- Pricing
- Not confirmed in the provided pages
Alternative a kodwai
ByteAsk
ByteAsk is a terminal-first AI coding agent for C and C++ that edits repositories and verifies changes with the real compiler, debugger, sanitizers, and tests before showing a diff. It offers a free tier plus paid plans, with editor connectors and zero-retention handling described in the source.
Manta AI
Manta AI is an autonomous web app testing tool for teams that want to map application behavior, catch regressions, and generate tests without writing scripts or maintaining selectors. It works from a URL and supports plain-English test flows, run results with screenshots, and scheduled or deployment-triggered checks.
CreateOS Sandbox
CreateOS Sandbox is an isolated compute environment for running code and agent workloads inside Firecracker micro-VMs. It is designed for workflows that need machine-level isolation, private networking between sandboxes, and programmatic control through SDK, CLI, or MCP.
hob
hob is an independent workspace for coding agents that keeps agent sessions, terminals, history, and follow-up work organized around the tools and providers you already use. It is aimed at developers who want local control over routing, history, and workspace structure rather than a bundled model stack.
SonOf
SonOf connects to your repo and PM tool, audits the codebase and surrounding product context, and turns approved work into shipped tickets with senior engineering review. It is aimed at founders and engineering leaders who need backlog help without hiring a full team immediately.
Ghost
Ghost è un assistente AI da terminale per chattare, generare codice ed eseguire task da riga di comando. Include modelli gratuiti, supporta Linux, macOS e Windows, ed è open source.