Skill UI

This skill provides a structured framework for managing complex, multi-agent development tasks. It treats agents as a collaborative team, utilizing structured work items, defined ownership, Agent Kanban boards, and explicit merge gates. This ensures that large-scale, multi-tool, or cross-branch development efforts are predictable, highly visible, and result in reliably mergeable, production-ready artifacts.

An interactive tool designed to discover, select, and assemble specialized teams of AI agents. It supports structured agent collections across different domains (e.g., Engineering, Marketing). Users select multiple roles for a specific task, and the skill launches them in parallel, synthesizing their independent findings into a unified report, including consensus and conflict analysis.

This skill guides users through evidence-based terminal operations. It is designed for scenarios requiring direct command execution, Git status checks, debugging CI/build failures, and pushing narrow, verified fixes. It emphasizes following a structured, minimal-scope workflow and accurately reporting the repository's current state (e.g., local changes vs. pushed code).

This skill intercepts the AI response flow to allow users to explicitly control the desired depth, length, and token usage of the generated answer before the model begins writing. By analyzing the prompt's complexity and the user's intent (e.g., 'short version', 'detailed answer'), it estimates the appropriate token range and prompts the user to select a granularity level (25%, 50%, 75%, or 100%).

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 tool automates the process of optimizing investment portfolio allocations using a mean-variance engine. It integrates neural prediction models and risk constraints to determine optimal asset weights. Users can set specific risk targets, assess comprehensive risk metrics (variance, correlation), and generate actionable, step-by-step rebalancing plans for professional portfolio management.

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.

Tree Ring Memory is a local-first, framework-agnostic layer for managing durable knowledge in AI agents. It provides structured methods for recalling past decisions, recording evidence, auditing project state, and deliberately forgetting information, ensuring the core project memory remains clean and actionable without relying on raw conversation transcripts.

This powerful tool utilizes Playwright to record high-quality, professional web application demo videos. It simulates natural user interactions, injecting a visible cursor overlay and adhering to specific pacing guidelines. It is ideal for creating comprehensive tutorials, feature walk-throughs, or documentation for stakeholders.