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.
How to fill out the California Small Claims Court SC-100 form, including field mapping and required vs. optional fields.
Run DBSCAN clustering with a custom distance metric using sklearn, including how to define weighted Euclidean metrics and extract cluster centroids.
Parallelize hyperparameter grid searches using joblib for CPU-bound tasks like DBSCAN clustering evaluation loops.
Compute Pareto-optimal frontiers from multi-objective optimization results, including dominance checking and frontier extraction.
Extract CVSS (Common Vulnerability Scoring System) scores from vulnerability data sources with proper fallback handling. This skill covers understanding CVSS v3, handling multiple score sources (NVD, GHSA, RedHat), implementing source priority logic, and dealing with missing scores in security reporting.
Use Trivy vulnerability scanner in offline mode to discover security vulnerabilities in dependency files without internet access.
Generate structured CSV security audit reports from Trivy JSON vulnerability data with severity filtering and proper field mapping.
Calculate distances between geospatial points and boundaries using GeoPandas with proper metric projections (EPSG:4087).
Work with PB2002 plate boundary and polygon data to identify plate membership and boundary distances for earthquakes.
Parse and work with USGS earthquake GeoJSON data, extracting IDs, magnitudes, coordinates, times, and place descriptions.
How to extract and format enterprise product data for specific queries. Use this skill whenever the user asks to retrieve employee IDs, report authors, competitor URLs, or insights from JSON enterprise product or metadata files.
Python patterns for analyzing large JSON datasets to find specific information, track tokens, and write answers in the required format.
Fuzzy search SEC 13-F COVERPAGE.tsv to find fund names and accession numbers by approximate name matching.
Compare fund holdings between two quarters (Q2 vs Q3) to identify increased/decreased positions by dollar value or share count.
Parse and analyze SEC 13-F filing TSV datasets to extract AUM, holdings count, and fund details by accession number.
Security patching for Apache Druid - covers JavaScript execution vulnerabilities, sampler endpoint protection, and filter validation patterns.
Security considerations for Jackson @JacksonInject - preventing JSON input from overriding injected values, covering CVE patterns and defense strategies.
Workflow for creating and applying patches to Java projects - diff format, patch application, and Maven rebuilds for security fixes.
How to use `gh api` with pagination to retrieve all results from GitHub REST API endpoints, including PRs and issues.
Analyze GitHub issues for a date range, counting totals, bug reports (by label substring), and closed bug reports.
Compute PR statistics (total, merged, closed, avg merge time, top contributor) from GitHub API data using Python or jq.
Guide for implementing preference optimization methods (DPO, SimPO, IPO) in PyTorch. Use when implementing loss functions for RLHF-style training.
Implements the SimPO (Simple Preference Optimization) loss function from the paper "SimPO: Simple Preference Optimization with a Reference-Free Reward". Use this when implementing or understanding SimPO training objectives.
Setup and installation of TRL (Transformer Reinforcement Learning) library with compatible torch/transformers versions. Use when setting up preference optimization training environments.
Handle {{IF_CONDITION}}...{{END_IF_CONDITION}} conditional blocks in Word templates — keep content and strip markers when true, remove entire block when false.
Robustly replace {{PLACEHOLDER}} tokens in Word documents, handling split runs across paragraphs, tables, headers, and footers.
Manipulate Word documents (.docx) using Python, including placeholder replacement and conditional section handling.
Look up arxiv paper metadata (title, abstract, subject) from arxiv IDs to classify papers by topic without reading full PDFs.
Organize files into subject folders using keyword-based classification of titles and abstracts, with fallback to full text extraction.
Use this skill to extract text from PDF files to determine their subject matter.
Implementing immutable fluent builder pattern in Scala as a translation of Python classes with method chaining, emphasizing immutability and type safety.
Using the circe library for JSON encoding, decoding, and manipulation in Scala as a replacement for Python's json module, covering the Json ADT, parsing, and path traversal.
Translating Python Enum classes to Scala sealed abstract classes or sealed traits with companion objects, preserving value fields and exhaustive pattern matching.
Parse meeting request emails to extract duration and time constraints, then find the earliest available time slot on a calendar.
Read input JSON files with email requests and write results.json with reply filenames and recipients.
Use PyMuPDF (fitz) to extract visual calendar blocks from a PDF, including color detection and pixel-to-time conversion.
Add a categorical color legend to a D3 chart using colored rectangles or circles with text labels, supporting both SVG-inline and HTML overlay styles.
Build a force-simulation bubble chart in D3.js v6 where bubbles are sized by a numeric value, colored by category, and clustered by category using forceX/forceY. Covers deterministic layout, collision, and tick-based settling.
Build a sortable, highlightable HTML data table linked to a D3 chart so clicking a chart element highlights the matching table row and vice versa.
Calibrating GLM parameters (Kw, coef_mix_hyp, wind_factor, lw_factor, ch) to minimize RMSE against field observations.
Running the General Lake Model (GLM3) for lake temperature simulation, including configuration file structure and execution.
Extracting and analyzing GLM NetCDF output to compute RMSE against field observations using exact datetime+depth matching.
How to structure and write a valid travel itinerary JSON file matching the required schema for /app/output/itinerary.json.
How to query and filter travel itinerary datasets (restaurants, accommodations, attractions, distances) to build constraint-satisfying travel plans.
Extract I-frame keyframes from video files using FFmpeg command line, saving them as numbered PNG images.
Count occurrences of an object in an image using OpenCV template matching (cv2.matchTemplate). Use this skill whenever the user needs to detect and count how many times a small reference image (template) appears in a larger image, such as counting coins, enemies, or other game sprites. Works on both grayscale and color images.
Convert images to grayscale in-place using Pillow (PIL), overwriting original RGB files.
Using INDEX&MATCH (single and dual condition) for dynamic lookups across rows and columns in Excel, including cross-sheet references.
Editing Excel files with openpyxl to insert formulas while preserving all existing formatting, styles, colors, and structure.
Calculating GDP-weighted means and other weighted statistics in Excel using SUMPRODUCT, including net exports as percent of GDP.
Guidelines for applying Anthropic's visual identity and brand tokens in technical and marketing artifacts.
Technical image generation using Python's Pillow library. Use this when you need to programmatically create diagrams, posters, or technical drawings.
Rules for rhyming in poetry based on modern Mandarin (Pinyin).
Rules for composing Chinese seven-character regulated verse (Qi Yan Lv Shi).
Methods for programmatically filling PDF forms (AcroForms) and saving the results using pypdf.
Techniques for inspecting PDF form fields and their properties using Python libraries like pypdf.
Implementation of a custom distance metric for DBSCAN clustering using scipy and sklearn.
Greedy matching algorithm to pair predicted cluster centroids with ground truth points for F1 and distance metrics.
> How to parallelize hyperparameter grid search using joblib. Use this skill whenever the user needs to evaluate many hyperparameter combinations (e.g., DBSCAN epsilon, min_samples, shape_weight) efficiently across CPU cores. Covers joblib.Parallel, parameter grid generation with itertools, and collecting results into a DataFrame.
Identify Pareto-optimal points from a set of multi-objective solutions.
Claude Skills — questions
Answers built from the skills we actually parsed.