Ask clarifying questions and get approval before implementation.
Review release readiness by checking ChangeLogs and version definitions and then create a git tag for the requested version. Use when asked to release or tag a version (e.g., "Please release 0.2.0", "Please tag 0.2.0") or to verify ChangeLog/version consistency before tagging.
Close task in tracker. Use when the user invokes /close or asks to close a tracker issue.
Build type-safe global state in React with Zustand. Supports TypeScript, persist middleware, devtools, slices pattern, and Next.js SSR with hydration handling. Prevents 6 documented errors. Use when setting up React state, migrating from Redux/Context, or troubleshooting hydration errors, TypeScript inference, infinite render loops, or persist race conditions.
Systematic bug investigation and root cause analysis
Use when partner provides a complete implementation plan to execute in controlled batches with review checkpoints - loads plan, reviews critically, executes tasks in batches, reports for review between batches
Update repository AGENTS.md for Option B (Tauri v2 + React/TypeScript desktop local-first) with executable commands, guardrails, and security/permission rules; creates backups and can add nested AGENTS.md.
Three.js asset loading - GLTF, textures, images, models, async patterns. Use when loading 3D models, textures, HDR environments, or managing loading progress.
MUST BE USED for code review. Use PROACTIVELY when /review command is invoked or after any code changes to check correctness, architecture, and security.
Production-ready reinforcement learning algorithms (PPO, SAC, DQN, TD3, DDPG, A2C) with scikit-learn-like API. Use for standard RL experiments, quick prototyping, and well-documented algorithm implementations. Best for single-agent RL with Gymnasium environments. For high-performance parallel training, multi-agent systems, or custom vectorized environments, use pufferlib instead.
Use when improving run configuration, seeding, and reproducibility for Python research code.
Use when adding formatting, linting, and type checking to Python research code.
Find, install, and configure an MCP server for a specific service. Handles discovery, installation, and verification.
Investigate a codebase question using an isolated subagent. Use when you need to read many files without bloating the main context.
Set up a build/test/lint validation loop for the current project. Detects the tech stack and configures CLAUDE.md with verification commands.
Audit and improve CLAUDE.md files. Use when rules aren't working, context is bloated, or after learning a lesson that should become a rule.
Discover and install useful Claude Code plugins for your project. Recommends based on your tech stack.
Plan a new feature before writing code. Explores the codebase, creates a structured plan with verification criteria, and only implements after approval.
Convert a YouTube video transcript into Claude Code skills. Extracts tips, evaluates against selection criteria, and generates SKILL.md files with full transparency.
This skill should be used when the user asks to "initialize a yFiles application", "create a new yFiles demo", "scaffold a yFiles project", "set up GraphComponent", or mentions "create app", "new demo", "initialize yFiles", "setup yFiles", or "License.value".