Skill UI

This utility detects and redacts sensitive personally identifiable information (PII) from screenshots and images. It can mask various data types, including IDs, credit card numbers, API keys, phone numbers, and emails. Crucially, all processing is conducted locally and offline, ensuring that sensitive data never leaves the user's machine. It is essential for secure debugging, data sharing, and analysis workflows.

Transform raw structured data (CSV, Excel, JSON) into professional, long-form, interactive visualization reports. This tool generates comprehensive business review pages featuring KPI dashboards, multiple dynamic charts (using Chart.js/ECharts), detailed data tables, and narrative business insights, making it ideal for weekly or monthly performance reporting.

This skill allows users to extract structured data (text, tables, attributes) from rendered web pages. It enhances traditional web scraping by ensuring all extracted strings pass mandatory AIDefence PII checks and prompt-injection gates before being stored or returned to the model. It supports building reusable extraction templates and running one-shot queries.

This utility detects and classifies various anomalies (such as spikes, flatlines, drift, or pattern breaks) within telemetry data streamed from IoT devices. It utilizes Z-score anomaly detection to provide actionable insights. This is crucial for investigating unusual device metrics, ensuring system health before deploying major firmware updates (canary advancement), or triaging fleet-wide operational alerts.

This memory bridge ingests Claude Code's local auto-memory markdown files and structures them into AgentDB. It generates high-fidelity ONNX embeddings, handles deduplication using cosine similarity, and indexes the memories in a scalable namespace. This enables seamless, unified semantic search across all historical sessions and projects, significantly enhancing RAG capabilities.

This guide provides a comprehensive workflow for searching and retrieving biomedical literature from PubMed and NCBI. It covers advanced query construction using MeSH terms and specific field tags, advanced filtering (date, publication type), and a repeatable programmatic workflow using NCBI E-utilities (esearch, efetch). Essential for researchers conducting systematic reviews and literature monitoring.

This skill provides a structured framework for designing and implementing composable recommendation, ranking, and feed pipelines. It models the industry-standard six-stage pattern: Source, Hydrator, Filter, Scorer, Selector, and SideEffect. Use it for any system that selects "the top K items for a (user, context)", such as social feeds, ad ranking, RAG rerankers, or content recommendation engines, ensuring a robust and scalable architecture.

This skill is designed for complex, iterative decision-making processes requiring deep computation, such as stochastic optimization, high-dimensional search, or repeated model rollouts. It mandates the creation of a structured 'Ledger' to rigorously record every trial, prior winner, fresh information, and decision mark. It ensures that conclusions are based on durable evidence, managing risk through explicit promotion gates and preventing premature certainty.

This is an advanced, structured workflow designed to conduct deep, multi-source investigations. It synthesizes current public evidence with user-provided context to generate highly structured reports. The workflow explicitly differentiates between verifiable facts, user-supplied evidence, logical inferences, and actionable recommendations, making it ideal for complex comparative analysis or continuous monitoring tasks.

This tool synthesizes multiple, scattered research findings retrieved from memory into a single, structured, and actionable report. It automatically grades evidence quality (High/Medium/Low), resolves conflicting claims, and provides concrete, recommendation-based insights, making complex data digestible for decision-making.

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 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.