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.
Procedures for setting up the python environment, installing dependencies, and verifying installations.
Guidelines for implementing the SimPO loss function as per the SimPO paper.
Manipulate Word documents using python-docx, including text replacement and handling conditional sections.
Efficiently read and parse JSON files for use in automation tasks.
This skill covers organizing files into subdirectories by subject matter.
Modeling domain entities and transformations using case classes, companion objects, and type-safe hierarchies.
Idiomatic functional programming in Scala using sealed traits, pattern matching, and Option/Either for robust error handling.
Logic for calculating meeting slots based on constraints and availability.
Parsing JSON input files for structured data extraction.
Extracting text and metadata from PDF documents.
Efficiently binding data to DOM elements and handling user interactions like hover and click in D3.js.
Creating bubble clusters in D3.js using force simulations.
Creating data-driven HTML tables using D3.js.
This skill covers computing RMSE metrics from observational and simulated water temperature data.
This skill covers running the General Lake Model (GLM) and modifying its configuration file (glm3.nml).
Provides methods to search, parse, and filter local CSV files in /app/data/.
Handles the structural requirements and formatting of the 7-day travel itinerary JSON.
Use ffmpeg to extract key frames (I-frames) from a video file into a directory.
Use ImageMagick to convert an image to grayscale inplace.
Detect and count occurrences of template images within a larger target image.
Advanced data retrieval in Excel using INDEX and MATCH.
Techniques for statistical analysis in Excel including weighted means, percentiles, and rounding.
Create clean technical posters and diagrams using the Python Pillow library, including shape drawing, typography placement, and annotation.
Explains the rules, forms, and modern rhyming conventions for composing a seven-character regulated verse (Qi Lu).
How to write multi-line text files using command line interfaces.
Guide on using the Python pypdf library to read, write, and fill PDF forms programmatically.
Custom distance metric implementation for use with clustering algorithms like DBSCAN.
Parallel processing with joblib to speed up computationally intensive parameter grid searches.
Output vulnerability reports as a formatted CSV file.
Extract CVSS scores from Trivy JSON output.
Use Trivy for offline vulnerability scanning of dependency files like package-lock.json.
How to calculate distance between points and other geometries (like lines or polygons) in metric units using GeoPandas.
How to determine which points fall inside a specific polygon using GeoPandas.
Provides techniques for parsing and extracting data from JSON files in Python.
Techniques to find specific data points in nested Python dictionaries using recursion.
A skill for parsing and analyzing SEC Form 13F TSV files, including COVERPAGE, INFOTABLE, and SUMMARYPAGE, using pandas.
A skill for finding best string matches in datasets using fuzzy matching libraries like fuzzywuzzy or rapidfuzz.
Security and Jackson deserialization issues in Apache Druid, focusing on property bypasses.
Fetch repository metrics, issues, and pull requests over a specific date range using the GitHub Search API.
Parse datetime strings and compute time durations/averages for datasets (like PR merge times) in Python.
Using Hugging Face TRL library for building custom trainers for Preference Optimization methods like DPO, SimPO, etc.
Implement the SimPO (Simple Preference Optimization) loss function for LLM alignment. Use this skill whenever implementing SimPO training objectives, reference-free reward optimization, or Bradley-Terry preference loss with target reward margin. Triggers on: SimPO, preference optimization loss, average log probability reward, gamma margin.
A skill for replacing placeholders and handling conditional blocks in python-docx.
Extracts text from PDF, DOCX, and PPTX files using Python.
Classifies text into predefined categories using keyword matching heuristics.
Guide for using Circe in Scala 2.13 for parsing and generating JSON, specifically when translating Python JSON handling.
Guide for creating functional fluent builders in Scala.
Guide for translating Python TypeVars and generics to Scala covariant and contravariant types.
Guide for translating Python @overload and dynamic typing (Any, Union) to Scala method overloading and type matching.
Extracts vector graphics, colors, and text from PDF files using PyMuPDF to analyze calendars or visual schedules.
Uses regular expressions to parse dates, time ranges, and durations from natural language text like meeting requests.
Guide for creating clustered bubble charts using D3.js force simulations (v6+), including collision and categorization.
Guide for linking D3.js SVG visualizations with DOM elements like HTML tables for bidirectional interactivity (hover, click, highlight).
Calibration guidance for GLM tasks. Often most effective after glm-basics has clarified the setup; glm-output is the companion skill for exact final metric computation.
Matching observations with simulated data using exact datetime and rounded depth merges in pandas.
Reading and processing NetCDF outputs from models like GLM using Python and xarray.
Using Python and Pandas to analyze large CSV datasets and filter based on specific criteria like budget, location, and amenities.
Algorithmic planning of itineraries avoiding specific transport modes and allocating time based on budget constraints.
Convert images to grayscale in-place.
Extract key frames (I-frames) from video files using FFmpeg.
Answers built from the skills we actually parsed.