Skill UI

Analyzes Excel and CSV uploads with DuckDB, inspecting schemas, running SQL, summarizing statistics, and exporting results so users can explore structured data efficiently.

This expert tool guides users through rigorous statistical analysis, specializing in A/B testing, hypothesis testing, and interpreting experimental results. It helps define sample sizes, executes appropriate tests (Z-test, T-test, Chi-square), and critically distinguishes between 'statistical significance' and 'practical significance' to ensure data-driven decision-making for product and business strategy.

This skill provides comprehensive commands to manage the entire AI model evaluation lifecycle using Arize. Users can list, retrieve metadata, and export results from experiments. It is ideal for comparing multiple model runs, benchmarking performance against a dataset, and analyzing model accuracy using structured evaluation workflows (CRUD operations).

This skill guides the systematic process of identifying and verifying candidate football entities (players or clubs). Use it when initial search results suggest a lead, or when the question provides enough detailed constraints (e.g., nationality, era, league) to hypothesize a candidate. The technique involves constructing targeted name-based searches and cross-checking every constraint (birth year, career timeline, etc.) against the found profile to ensure accuracy.

This advanced research technique is designed for deep dives into football topics. Instead of relying solely on the original query, it mandates tracking every discovered entity—such as player names, club names, or coaches—from search results. By treating these entities as new leads, the user must formulate follow-up searches using the discovered names and remaining constraints, ensuring nothing relevant is overlooked and building a comprehensive knowledge chain.

This technique guides users on how to construct highly specific search queries for academic papers. Instead of relying solely on broad topic keywords, users should prioritize combining concrete numerical identifiers—such as exact sample sizes, population percentages, concentration levels, or specific year ranges—with methodological terms and discipline keywords. Using these numerical anchors significantly increases the uniqueness and precision of the search results, helping to pinpoint the exact study required.

A powerful tool designed for academic researchers and scientists. It searches a curated database of scientific literature and retrieves structured metadata and quantitative experimental data (including methods, results, sample sizes, and quality scores) directly from full-text papers, going far beyond simple abstracts. Ideal for systematic literature reviews, evidence synthesis, and meta-analyses requiring detailed, comparable study information.

Access and query 78 public databases covering diverse fields including biomedical, physical sciences, earth science, and economics. This skill identifies the relevant database(s) for user queries (e.g., genes, compounds, patents, climate data, financial indicators), executes structured REST API calls, and returns raw JSON results, providing a comprehensive data retrieval pipeline.

A comprehensive guide and toolkit for optimizing Hex platform costs. This resource details strategies for reducing compute expenditure by auditing seat licenses, consolidating data connections, and optimizing notebook run schedules. It includes code examples for implementing result caching and usage monitoring to prevent unexpected billing spikes.

This skill provides comprehensive access to Bright Data's SERP API and Web Scraper API. It allows users to collect structured search engine results (Google, Bing, Yahoo) synchronously, or trigger large-scale, asynchronous data collections with webhooks. Ideal for market research, SEO analysis, and large-scale data scraping projects.

This skill performs a rigorous, zero-context audit to verify that every quantitative claim, comparison, and scope statement made in a research paper precisely matches the underlying raw experimental results. It is designed to eliminate confirmation bias by treating the paper text and raw data as independent evidence sources.

A comprehensive guide to diagnosing and enhancing search relevance within Qdrant. This skill addresses reported issues such as low precision, low recall, or irrelevant matches. It guides users through determining if the problem lies in the embedding model, Qdrant configuration, or the query strategy. Techniques covered include optimizing chunk splitting, tuning HNSW parameters, implementing hybrid search, and utilizing reranking for superior result quality.