Skill UI

This skill provides an interactive, guided tutorial for beginners learning GitHub Copilot CLI. It features separate learning tracks for both developers and non-developers, covering everything from basic terminal usage to advanced planning commands. Use this skill to make the command line approachable and fun, regardless of your technical background.

A comprehensive skill for capturing and documenting web application and graphical UI states during the development cycle. It supports full-page captures using Playwright, reliable post-processing cropping with PIL, capturing interactive states (hovers, selections), and even handles non-web desktop applications using mss/ctypes.

This skill provides a generic entry point for delegating self-contained, complex tasks to various LLM-powered sub-agents (Codex, Claude Code, Pi). It is designed to be used in meta-skill DAG steps for tasks like feature creation, PR reviews, code refactoring, or document generation, providing full tool surface access. It favors non-interactive CLI modes for reliable background execution.

A comprehensive guide for authors preparing the supplementary material for multimedia conferences (e.g., ACM MM). It details how to structure the content, separating core arguments (Body) from supporting evidence (Supplement). Learn best practices for packaging videos, audio, and interactive demos, ensuring technical compatibility (H.264/MP4), maintaining anonymity for double-blind review, and creating a legible, reviewer-friendly submission index.

A robust, shell-free utility designed to launch executables and their arguments into a visible host terminal window. It supports multiple advanced modes including dry-run inspection, capability detection, detached fallback, and standalone recovery. Ideal for applications needing to securely execute interactive CLIs, SSH sessions, or local development processes in a reliable, non-interpolated manner.

This skill allows delegating self-contained coding tasks to assistants like Codex or Claude Code via non-interactive background processes. It supports feature development, code review, and refactoring. Use OpenSquilla's exec_command and process tools to manage sessions, poll output, and send inputs without blocking the main workflow. Ideal for tasks requiring substantial file exploration.

This skill enables non-interactive usage of GitHub Copilot CLI for code reviews, questions, and snippet generation. It requires explicit user consent before sending local files or context to ensure privacy. Use for secondary advice without manual interaction.

Runs Claude Code non-interactively for code analysis and implementation tasks. Use when delegating repository work to Claude Code or comparing results with another coding agent. Ensures read-only access by default and requires explicit permission for modifications.

Runs Codex CLI non-interactively for code analysis and implementation tasks. Delegates repository work to Codex and obtains coding-agent results via JSONL output. Supports sandbox modes and authentication checks.

Runs GitHub Copilot CLI programmatically for repository analysis and coding tasks, with non-interactive prompts, JSONL output parsing, and secure tool permissions.

This skill automates Kimi Code execution in non-interactive mode within a bounded repository. It parses event streams, handles authentication securely, and manages sandbox permissions. Use it to delegate specific coding tasks to Kimi Code while ensuring safety constraints like timeouts and directory limits. It prevents interactive UI launches and avoids credential exposure.

This skill executes the OpenClaw local agent non-interactively to delegate bounded tasks. It sets up the working directory, checks for the openclaw command, and runs the main agent with a JSON response. It handles authentication, permissions, and ensures read-only access unless workspace changes are requested. It parses JSON payloads even on success to detect failures.