Skill UI

A comprehensive pattern for developing board-based puzzle games (e.g., Match-3, Sokoban, Logic Grids). This skill focuses on modeling the board state, move input validation, and the core rule resolution loop. It defines how cascades, scoring, and undo functionality must operate deterministically to ensure fair and bug-free gameplay, separating game logic from visual rendering.

Structure a pygame (pygame-ce) game in Python: init/event/update/draw loop, delta-time movement, Surface/Rect blitting, keyboard/mouse input, and Sprite/Group management with collision. Use when building or debugging a pygame game.

Use this skill to formally report bugs, file issues, or submit problems encountered with the canvas-apps plugin. It facilitates the submission of detailed bug reports directly to the associated GitHub repository, helping developers track and resolve technical problems.

Script Roblox experiences in Luau: get services, create and parent Instances, connect events, run server Scripts vs client LocalScripts, and communicate across the client/server boundary with RemoteEvents and RemoteFunctions under a server-authoritative model. Use when building or debugging Roblox Studio scripts.

A comprehensive knowledge base covering the theoretical fundamentals and advanced algorithms of Deep Reinforcement Learning (Deep RL). It details core concepts like Bellman equations, policy gradients, and the trade-offs between policy optimization and Q-learning. The guide analyzes leading algorithms such as PPO, SAC, TRPO, and DDPG, and provides rigorous best practices for debugging and running scientifically sound RL experiments.

A disciplined methodology for fixing bugs and small behavioral changes with minimal impact. It involves tracing symptoms to the narrowest responsible mechanism, implementing changes only at that layer, and adding highly focused regression proofs to ensure surrounding functionality remains intact. Ideal for critical systems requiring maximum stability.

A safe, cross-platform copilot for all command-line interface (CLI) needs. It detects the user's environment (OS, shell, installed tools) before suggesting or running commands. It excels at handling complex, long-running sessions (REPLs, SSH, databases) and debugging common errors like 'command not found' or 'permission denied' across Windows, macOS, and Linux.

This skill covers setting up a foundational Three.js scene. It includes configuring import maps, initializing the Scene, PerspectiveCamera, and WebGLRenderer trio, establishing a responsive render loop, and adding OrbitControls. Use this when bootstrapping a new Three.js application or debugging rendering issues like blank canvases. It excludes model loading and material lighting setup.

Propose labels and categorize GitHub issues for maintainers without direct edits. Checks duplicates, verifies bug info, and drafts comments based on repository rules. Optimizes issue management workflow for googleapis/mcp-toolbox.

Skill for parent orchestrators to delegate independent workstreams to ephemeral context workers via delegate_task. Covers mapping needs, parallel unit writing, high-risk changes, adversarial review, and multi-step debugging. Enforces named-reason delegation, self-contained missions, and verification of worker output.

This agent is dedicated to systematically exploring technical ideas, such as new features, refactoring, or bug fixes, *before* committing to code changes. It investigates the existing codebase, analyzes current architecture, compares multiple potential approaches (including pros, cons, and effort), and generates a structured 'exploration' report with a clear recommendation and identified risks. It ensures thorough planning and risk mitigation during the design phase.

A comprehensive, black-box workflow designed for professional bug bounty hunting and penetration testing. It guides users through a 5-stage methodology (Intake, Recon, Enum, Hunt, Report), covering 19+ advanced attack playbooks (e.g., RCE, SQLi, XSS, SSRF). The skill integrates real-world exploit cases, structured payloads, and WAF bypass techniques for maximizing vulnerability discovery in web applications and APIs.