Skill UI

This skill allows users to generate complete songs, including vocals or pure instrumental tracks, using advanced AI APIs. By providing a structured JSON specification, you can define the desired style, mood, scene, and optional lyrics. Ideal for creating background music, theme songs, jingles, or full compositions for creative projects.

A powerful skill for generating or editing single images using external APIs via OpenRouter (defaulting to Gemini). Users can provide a detailed text prompt, specific aspect ratios, and image sizes (1K, 2K, 4K). It supports image-to-image editing by accepting an optional input image, making it ideal for concept art, product rendering, or modifying existing visual assets.

This specialized tool drafts a single, academically rigorous section of a research paper. It accepts detailed structured inputs, including a complete writing plan, overall paper outline, citation budget, and optional figure/table context. The output is a pure LaTeX fragment that adheres to strict structural conventions for various sections (e.g., Introduction, Method, Results). It ensures the section advances the paper's single narrative thread while maintaining professional academic quality.

This skill generates high-quality, short cinematic video clips (3-15 seconds) using the powerful Seedance 2.0 model. It accepts structured English prompts, an optional starting frame image, and identity/style reference images to ensure character and aesthetic consistency across shots. Supports multiple backends (OpenRouter, Volcengine, BytePlus) for flexible and reliable video generation.

This tool renders a static PNG image suitable for title screens, opening cards, or ending credits. It allows users to generate a centered headline and an optional subtitle on a solid-color background. Utilizing the Pillow library in Python, it supports CJK-friendly font fallbacks and is designed for deterministic, offline use in video meta-pipelines, ensuring reliable card generation without external dependencies or LLMs.

This powerful utility concatenates multiple numbered MP4 segments (e.g., 1_*.mp4, 2_*.mp4) into a single, cohesive video file. It supports optional fade transitions and normalizes video parameters, including resolution, FPS, and codec, ensuring a professional and unified output. It is ideal for stitching together short clips produced by automated workflows into a final reel or presentation.

A crucial post-hoc review skill designed for AI-generated Markdown prose. It audits content against a comprehensive set of 21 stylistic and structural rules using both deterministic mechanical checks and deep semantic judgment. Users can audit a single file to get a detailed scorecard, or compare two drafts (A/B) for regression testing. Furthermore, it can optionally generate a polished, revised copy in a separate file, ensuring the integrity of the original document while significantly improving linguistic quality and adherence to style guides. Ideal for refining AI-written documentation, ...

This comprehensive guide outlines modern and idiomatic Java best practices. It covers advanced topics like Streams, Optional, Records, and Pattern Matching, while detailing anti-patterns and efficient ways to handle concurrency and error management in contemporary Java development.

A comprehensive guide covering modern Swift best practices. Learn how to handle optionals safely, utilize functional transforms, manage concurrency with async/await, and adopt Protocol-Oriented Programming (POP). This reference helps developers write highly performant, idiomatic, and safe Swift applications.

The `hf-mem` skill estimates the necessary VRAM or system memory required to load Safetensors or GGUF model weights for inference directly from the Hugging Face Hub. It uses efficient HTTP Range requests, avoiding local downloads. Users can determine if a model fits on their GPU or instance, including optional estimations for the KV cache and setting limits like maximum sequence length.

A systematic skill designed to investigate data incidents, including freshness delays, volume anomalies, schema changes, and ETL failures. It guides the agent through a comprehensive process using Monte Carlo's observability metadata and optional direct data querying to pinpoint the root cause.

This skill provides comprehensive code review capabilities for analyzing git diffs or specific file scopes. It goes beyond simple syntax checks, assessing code for reuse opportunities, quality issues (e.g., redundant state, leaky abstractions), efficiency bottlenecks, and adherence to standards. It utilizes multiple specialized sub-agents (Reuse, Quality, Efficiency, Clarity) to perform parallel, read-only audits. Optionally, it can apply high-confidence, behavior-preserving fixes upon user request.