Skill UI

This skill executes comprehensive historical backtesting using the high-performance neural-trader engine. It supports features like walk-forward validation and captures key financial metrics (Sharpe, Drawdown). Critically, it signs the resulting artifact using Ed25519 cryptography, providing non-repudiable tamper evidence required for promoting research models to live trading systems.

This skill dispatches computationally intensive quantitative trading jobs (such as multi-year walk-forward analysis, large Monte-Carlo simulations, parameter sweeping, or deep model training) to a managed cloud agent runtime. It is designed for large-scale, production-grade backtesting that exceeds local machine capacity, ensuring robust execution and artifact management via Anthropic's cloud services.

This skill calculates model interpretability for algorithmic trading signals using PageRank. It constructs a feature-contribution graph from an LSTM or Transformer signal and executes a single-entry PageRank calculation. The result identifies the top-K features that most significantly drove the prediction, providing a "regulator-grade" audit trail essential for compliance with regulations like the EU AI Act.

A comprehensive tool for analyzing the current market regime (bull, bear, ranging, volatile) using advanced neural models. It detects the prevailing market structure, provides a confidence score, and recommends suitable trading strategies. Ideal for market analysis, backtesting preparation, or general market condition inquiries.

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.

A comprehensive guide to positioning research contributions for top-tier financial microstructure journals, such as the Journal of Financial Markets (JFM). It teaches authors how to articulate their novel findings by anchoring them within established academic lineages (e.g., Glosten-Milgrom) and explicitly identifying the gap relative to 3-5 closest prior works. Focus must be placed on market mechanics and trading-process advances, rather than general financial outcomes.

This guide helps academics determine if their research contribution is fundamentally about market microstructure—such as order flow, liquidity, price impact, or trading mechanisms—making it suitable for the Journal of Financial Markets (JFM). It provides a rigorous 'mechanism question' test to distinguish between microstructural findings and broader corporate or asset pricing issues.

This guide helps authors determine if their research, particularly in market microstructure, liquidity, or high-frequency trading, is suitable for the Journal of Financial Markets (JFM). It provides detailed heuristics on topic fit, required evidence, structural guidelines, and common rejection triggers, ensuring submissions are optimally framed for specialist financial economics venues.

This skill guides the construction of robust, production-grade backtesting systems for financial trading strategies. It helps users systematically validate performance, avoid common biases (like look-ahead bias), and implement advanced techniques such as walk-forward analysis, ensuring reliable performance estimates before live deployment.

Longbridge offers a comprehensive integration for accessing real-time market data, charts, and fundamentals across HK, US, A-share, and SG markets. It supports portfolio analysis, options trading, and sector screening. The tool is trilingual (Simplified Chinese, Traditional Chinese, English) and works via CLI or MCP, ensuring secure read-only access to account information and market insights.