Skill UI

This guide details the entire lifecycle of building a robust anti-phishing security awareness program. It covers everything from initial phishing simulations and designing role-specific training modules to continuous deployment, metric tracking, and optimization. The goal is to improve human security posture, ensure compliance, and build a strong defense culture against social engineering threats.

This skill manages the entire lifecycle of long-running deep learning training and inference jobs deployed on remote, ephemeral computing environments (e.g., AutoDL, RunPod, bare SSH, Slurm, K8s). It provides critical capabilities like resilient checkpointing, managing spot instance preemption, debugging resource failures (OOM, NaN), and ensuring safe billing teardown, making it essential for professional ML Ops on rented resources.

A comprehensive tool for identifying poisoned training data, backdoored ML models, and supply-chain compromises. It leverages IBM's Adversarial Robustness Toolbox (ART) for activation clustering and spectral signature analysis, uses Cleanlab for label quality checks, and enforces integrity via weight-hash verification and safetensors. Essential for securing ML pipelines before training, deployment, or during supply chain audits.

A comprehensive guide for training or fine-tuning advanced sentence-transformer models. This skill assists in selecting the correct architecture—including Bi-encoders (SentenceTransformer), CrossEncoders (for reranking), and SparseEncoders—for various Natural Language Processing (NLP) tasks. Use cases include semantic retrieval, similarity matching, document clustering, and pair-wise classification.

This skill provides expert capabilities for training and fine-tuning transformer language models using the TRL (Transformers Reinforcement Learning) library. It supports state-of-the-art post-training methods, including Supervised Fine-Tuning (SFT), Direct Preference Optimization (DPO), Group Relative Policy Optimization (GRPO), and Reward Model training. Use this skill to align and customize foundation models for advanced tasks via CLI commands.

A comprehensive presentation template designed for educational workshops and training sessions. It features a warm paper aesthetic, persistent learning objectives on the left sidebar, core content slides, interactive Multiple Choice Question (MCQ) self-assessment pages, and module wrap-ups, ensuring a structured and engaging learning experience.

This skill fetches the latest, live documentation and code examples for specific libraries and frameworks (e.g., Next.js, React, Prisma). Instead of relying on static or outdated training data, it utilizes the Context7 MCP to provide accurate, up-to-date answers for setup questions, API references, and code generation based on current best practices.

Provides a structured workflow for machine learning engineering tasks, covering model development, training, evaluation, and deployment. Use it to manage end-to-end ML projects efficiently.

This tool manages the lifecycle of learned intelligence patterns, implementing the crucial DISTILL and CONSOLIDATE phases of the 4-step intelligence pipeline. It enables the capture of successful task knowledge, supports multi-domain adaptation via SONA or MicroLoRA, and consolidates learned patterns into long-term storage to prevent catastrophic forgetting.

A skill covering common PyTorch patterns for building, training, and debugging deep learning models. Note: this file is a stub pending Japanese translation; refer to the original English skill file for full content.

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.

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.