Skill UI

Calculates statistical power and required sample sizes for study planning. It supports both closed-form methods (for standard tests like t-tests, ANOVA) and simulation-based methods (for complex models like mixed models or survival analysis). Use this skill to determine the minimum detectable effect (MDE) or justify sample sizes for grants and IRB protocols before data collection.

This skill provides a highly rigorous framework for analyzing complex public issues, platform governance, and institutional compliance. It is designed to structure evidence boundaries, assess procedural justice, and track public commitments by synthesizing source ledgers and multiple data points. It is intended as a specialized sub-routine triggered by the main CrossFrame suite, not for standalone use. Key focus areas include source verification, identifying weak signals, and mitigating AI compliance performance risks.

This skill guides the design of statistically rigorous A/B tests for optimizing product features, UI/UX changes, onboarding flows, and pricing models. It generates a complete test plan, including formulating hypotheses, calculating required sample sizes, estimating test duration, defining primary and guardrail metrics, and providing a detailed guide on interpreting results (ship, iterate, or reject).

This skill extracts precise, pixel-level data from various chart types (bar, line, scatter, pie) and graphs presented in images. It generates a comprehensive structured data table, including confidence levels, axis details, and assumptions, making the raw data ready for spreadsheets or further analysis. Ideal for digitizing charts, transcribing graph values, or turning visual data into actionable formats.

A structured framework for performing in-depth cohort analysis. It helps measure user retention rates, estimate Lifetime Value (LTV), and segment users based on behavior over time. Use this when analyzing growth metrics, identifying churn risks, or presenting product health to leadership.

This skill scrapes Substack Notes analytics by extracting key engagement metrics (likes, comments, restacks) and content details. It overcomes the lack of a public API, allowing users to analyze performance data from a specified date range into a highly formatted, actionable Excel spreadsheet. The output includes detailed per-note data, total engagement calculations, and summary statistics.

Run arbitrary PySpark or pure Python code directly on Fabric Spark compute using the Livy API. These sessions are ephemeral, meaning no notebook artifacts are created or persisted, making it ideal for automated ETL processes, programmatic data transformations, or agent-driven compute workflows. Provides full read/write access to lakehouse Delta tables via Spark SQL.

This guide demonstrates how to use the pycti Python client to model complex cyber threat intelligence within the OpenCTI platform. It structures threat actors, intrusion sets, campaigns, and TTPs into a unified STIX 2.1 knowledge graph. The skill is essential for correlating fragmented data from multiple sources (IOCs, reports) into a comprehensive, queryable adversary picture, supporting advanced threat hunting and detection engineering.

This comprehensive tool suite provides high-fidelity parsers for key Windows artifacts, including MFT, Registry hives, Prefetch files, and ShellBags. Designed for digital forensics and incident response, it converts raw system evidence into structured CSV/JSON formats. The output is optimized for feeding into Timeline Explorer, enabling analysts to build comprehensive, correlated timelines and establish a clear chain of events.

KAPE is a powerful, Windows-native triage tool used for Digital Forensics and Incident Response (DFIR). It executes in two phases: 'Targets' collect raw forensic artifacts (like registry hives, event logs, and prefetch files) from live systems, while 'Modules' parse this collected data into normalized, analyst-ready CSV/JSON formats. It enables fast, defensible, and scalable artifact collection, especially when full disk imaging is impractical.

A static analysis tool that audits SQL code for common cost and performance anti-patterns (e.g., full-table scans, Cartesian joins). It calculates a query health score (0-100) for data warehouses like BigQuery, Snowflake, and Redshift, providing a prioritized, actionable plan with concrete fixes and estimated cost savings. Essential for FinOps and data quality assurance.

This skill guides users through building, deploying, and maintaining machine learning applications on Hugging Face Spaces. It provides comprehensive knowledge on selecting the optimal SDK (Gradio, Docker, Static), choosing appropriate hardware resources (ZeroGPU, dedicated GPUs, CPU-basic), and following best practices to ensure ML applications are deployed efficiently and successfully on the platform.