Skill UI

Assess and manage investment risk using the neural-trader risk engine. This tool calculates critical metrics such as Value at Risk (VaR), Conditional VaR (CVaR), and Sharpe ratio for both single positions and entire portfolios. It also optimizes position sizing using methods like Kelly criterion and checks for market circuit breaker triggers (e.g., high correlation spikes or volatility regime shifts) to ensure robust trading decisions and minimize exposure.

This tool generates sophisticated trading signals by leveraging an advanced anomaly detection engine. It scans financial assets for various market anomalies—such as spikes, drifts, and oscillations—and classifies them. It then integrates neural network prediction and searches historical pattern databases to output highly ranked, actionable trading signals, complete with confidence scores and target parameters.

A comprehensive tool for training and evaluating advanced neural network models (including LSTM, Transformer, and N-BEATS) on time-series market data. It facilitates the prediction of financial trends, generating results with confidence intervals, and comparing performance across multiple architectures. Ideal for quantitative trading research and financial forecasting.

Provides a structured workflow for systematically accessing and analyzing official intellectual property data from the USPTO. This skill covers patent search (PatentSearch API), trademark status checks (TSDR), ownership transfer tracking, and building reproducible IP landscape logs. It is designed for data collection and record verification, emphasizing the use of official APIs and rigorous logging practices. (Warning: Not for legal advice.)

This skill helps discover natural groupings within a large collection of semantic embeddings using the `ruvector` tool. It employs graph-based algorithms (Spectral/Louvain) to cluster code or data, revealing underlying themes, identifying outliers, and organizing vast vector sets. Ideal for analyzing codebases or large datasets to understand relationships between entries.

This skill utilizes hyperbolic space embedding (Poincare ball model) to represent complex data structures with inherent hierarchy, such as dependency trees, ontologies, or module taxonomies. Unlike standard Euclidean embeddings, hyperbolic space preserves hierarchical distances efficiently, allowing for accurate distance calculations (geodesic distance) that grow logarithmically with tree depth. It is ideal for advanced knowledge graph analysis and code architecture mapping.

This plugin provides advanced vector search capabilities using two distinct paths. The large-scale path supports millions of vectors via HNSW, ideal for general corpus search. The hot-path router handles small, critical pattern routing (≤11 patterns) with sub-ms latency. Key features include RaBitQ 1-bit quantization for memory efficiency (32× reduction), hyperbolic space search for hierarchical data, and unified cross-namespace search.

A comprehensive skill for ingesting, analyzing, and manipulating multi-modal video and audio content. Supports real-time desktop capture, live stream monitoring (RTSP), and file uploads (URL/local). Capabilities include advanced indexing (visual, semantic, temporal), precise moment searching, professional timeline editing (subtitles, overlays, dubbing), transcoding, and generating actionable alerts. Ideal for content creators, monitoring systems, and complex media workflows.

This comprehensive tool automates the process of translating non-English official documents (images) for visa applications. It handles image preprocessing (conversion and rotation), performs robust OCR using multiple frameworks, translates the extracted text into professional English while maintaining document structure, and generates a certified bilingual PDF containing both the original image and the formatted translation.

A comprehensive skill to extract and transcribe subtitles from any YouTube video URL. It utilizes a two-stage approach: first, attempting fast extraction via the yt-dlp CLI tool; failing that, it falls back to advanced browser automation using Chrome DevTools MCP to scrape the transcript panel content directly from the webpage DOM. The output is always a clean, timestamped plain text file.