Claude Skills
The open format is called Agent Skills and works in Claude Code, Codex, Cursor and other agents — most people know it as Claude Skills.
Every Agent Skill we could find on GitHub, deduplicated by content. 79 600 files from 1 763 authors, of which 61 947 are unique — the rest is the same skill repackaged into someone else's repository. For each one: what it weighs in tokens, whether it ships runnable scripts, and which MCP servers it needs.
Efficient and robust Pareto frontier calculation for multi-objective optimization.
Greedy point matching and calculation of F1 score and average distance (delta).
Efficient DBSCAN implementation by caching distance matrices for repeated hyperparameter evaluations.
Structured approach to grid search with nested loops and parallelization.
Reliable CSV generation for security audits with automated field mapping.
Optimized Trivy offline scanning for lock files, including all dependency types.
Robust extraction of vulnerability metadata with CVSS v3 priority and v2 fallback.
Selecting optimal projections for distance calculations in the Pacific.
Techniques for handling geometries that cross the 180/-180 degree longitude line.
Advanced techniques for analyzing complex enterprise product JSON files containing messages, documents, and transcripts.
Precise identification of user roles and their contributions to product discussions and competitor analysis.
Advanced fund analysis with share change verification to distinguish between price movement and deliberate investment.
Professional fund search with accession number verification and report type checking.
Robust stock analysis and holder identification, bypassing file path issues in default scripts.
Optimized Maven build command for Apache Druid to verify security patches.
A checklist for ensuring thorough patching of Apache Druid security vulnerabilities.
Detailed explanation of the Jackson empty key ("") vulnerability and how it bypasses @JacksonInject security.
Mastering GitHub API interactions using curl, including search, pagination, and rate limit handling.
Robust Python logic for community pulse metrics with substring label matching and precise PR status handling.
Procedures for aligning a project-specific environment to its `environment.yml` and project-root structure.
Procedures for resolving environment conflicts, broken packages, and restricted system-wide access in Python.
Precision implementation of SimPO (Simple Preference Optimization) loss with length-normalized rewards and configurable margin.
Techniques for preparing and formatting data from JSON/API sources for inclusion in documents.
Advanced techniques for Word document automation, including handling split runs, headers/footers, and complex conditionals.
Advanced document classification using weighted keyword scores and expanded subject vocabularies.
Safely moves files and validates the integrity of the organization process.
Robustly extracts text from various document formats with multiple fallback options.
Robust temporal tokenization with Java 8 Time API and DateTimeFormatter.
Refined FP patterns for Functors and Monads in Scala, including applicative.
Improved JSON path navigation using Circe cursors in Scala.
Enhanced numeric tokenization with BigDecimal and precision in Scala.
Robust text tokenization with regex and position tracking in Scala.
Advanced PDF parsing for calendar extraction, including sidebar filtering and color-based classification.
Advanced scheduling with priority-based sorting (EDF) and flexible slot overwriting.
Advanced clustering techniques in D3.js force simulations for compact and aesthetically pleasing layouts.
Enhancing user experience with polished tooltips, table interactions, and smooth transitions.
Robust data cleaning and formatting for financial datasets in D3.js.
Advanced GLM configuration and calibration strategies for vertical temperature profiles.
Robust GLM evaluation methods for deep and summer water temperatures.
Advanced filtering for pet-friendly and cuisine-specific travel data.
Validate travel budgets using accommodation and restaurant data.
Plan 7-day itineraries with logical city transitions and transport.
Advanced FFmpeg techniques for keyframe extraction with quality control.
Efficient inplace image processing using ImageMagick's mogrify.
Robust object counting using template matching with parameter tuning.
Advanced two-condition lookups in Excel using INDEX and MATCH with sheet-specific references in openpyxl.
Using Excel's statistical functions with rounding and percentage scaling in openpyxl.
Verifying Excel formula results and handling errors using recalc.py and openpyxl.
Weighted mean calculations in Excel using SUMPRODUCT and SUM in openpyxl.
Refined skill for enforcing Anthropic corporate brand identity with precise HEX token usage.
Enhanced skill for programmatic design, incorporating system-level dependency management for image generation libraries.
Provides guidelines on modern Mandarin pronunciation and rhyming for classical forms.
Advanced guidance for writing regulated verse, covering parallelism, rhythm, and emotional depth.
Improved skill for filling XFA/non-standard PDFs using FreeText annotations when field mapping is unavailable.
Implementation of DBSCAN with custom weighted Euclidean metric, optimized for parallel execution.
Efficient greedy matching of cluster centroids to expert points with distance constraint.
Identifying the Pareto frontier for a set of results, efficient implementation.
Robust JSON-to-CSV parser for Trivy vulnerability reports.
Run Trivy in offline mode to identify high and critical vulnerabilities in package-lock.json with specific flag usage.
Geospatial data analysis with geopandas for distance calculations, filtering, and modern API usage.
Claude Skills — questions
Answers built from the skills we actually parsed.