发现优质的 AI Agent 技能

> **Official Documentation**: [docs.multiversx.com](https://docs.multiversx.com) > **Network Explorer**: [claws.network](https://claws.network)

> Define all data shapes BEFORE implementation. All agents reference this document.

You help users create and maintain a product marketing context document. This captures foundational positioning and messaging information that other marketing skills reference, so users don't repeat themselves.

**Vernox Utility Skill - Perfect for document digitization.**

This document describes how Value at Risk (VaR) and Conditional VaR (CVaR / Expected Shortfall) are calculated in RiskOfficer. All calculations run in the **RiskOfficer backend**; portfolio returns are built from position weights and historical prices (log-returns where applicable).

This document describes how **stress tests** are implemented in RiskOfficer. They use **historical crisis periods** and re-use the same portfolio valuation and drawdown logic as elsewhere.

This document describes how **Risk Parity** (Equal Risk Contribution, ERC) optimization is implemented in RiskOfficer. The logic runs in **ComputeService** (rebalancing) and is also used as the fallback for Max Sharpe in the construct-portfolio flow.

This document describes how the **pre-trade risk check** is implemented in RiskOfficer. It runs in the **RiskOfficer backend** (risk API); the endpoint is **synchronous** and **FREE** (no subscription required).

This document describes how **Monte Carlo** portfolio forecasts are implemented in RiskOfficer. The logic runs in the **RiskOfficer backend** (risk_service).

This document describes how **Sharpe ratio**, **volatility**, **max drawdown**, and related metrics are computed in RiskOfficer. They appear in VaR responses, optimization results, and construct-portfolio outputs.

This document describes how **Hierarchical Risk Parity (HRP)** is implemented in RiskOfficer for portfolio construction (auto-generate flow).

This document describes how **cross-portfolio PnL correlation** and optional **crisis regime analysis** are implemented in RiskOfficer. The logic runs in the **RiskOfficer backend** (Celery task `correlation.run_correlation_analysis`); market data (historical prices) comes from the Data Service.