Developer Tooling Skill Leaderboard

Make agent responses concise, structured, and action-first.

Find and assemble reusable agent workflows.

Optimize React applications and Vercel deployments.

Build applications and create documents and multimedia.

Structure brainstorming and agent development workflows.

Build interactive interfaces, games, and visual effects.

Develop Vue applications and maintain JavaScript tooling.

Test local code against live Kubernetes environments.

Add image understanding to text-only agents.

Plan projects and coordinate coding agents and tools.

Build AI, robotics, and accelerated computing workflows with NVIDIA tools.

Chat with coding agents through messaging platforms.

Integrate typed decisions and calibrated probabilities with TypeSafe.

Coordinate coding tasks from implementation to reviewed commits.

Design and audit reliable agent architectures.

Develop applications and automate document and media workflows.

Control Unreal Editor and build gameplay systems.

Build customizable Python agents with tools and memory.

Build agents that learn and reuse persistent skills.

Automate development, research, and content workflows.

Compare model API prices across providers.

Diagnose and repair Windows Codex Desktop features.

Build autonomous coding loops with verifiable development environments.

Benchmark and reduce Xcode build times.

Build, debug, and optimize Apple platform applications.

Coordinate coding agents through verified implementation workflows.

Configure coding agents for development and automation workflows.

Curate and manage reusable agent skill libraries.

Migrate Kotlin projects and improve build tooling.

Build websites, generate assets, and deploy services.

Control websites and desktop apps through OpenCLI.

Search and map codebases using syntax trees.

Guide users through macOS tasks with onscreen pointers.

Run team AI agents with tools and memory.

Plan and deliver customer engineering engagements.

Find and compare Jev and TypeSafe resources.

Build command-line interfaces with the Gunshi library.

Run code and tests in isolated E2B sandboxes.