Skill UI

Master the implementation of Firecrawl webhooks for robust, real-time handling of asynchronous crawling and batch scraping jobs. Instead of inefficiently polling the crawl status, configure webhook endpoints to receive instant POST events (e.g., `crawl.page`, `crawl.completed`). This guide covers event types, signature verification using HMAC-SHA256, and processing streamed data for building advanced data pipelines.

Ray Data offers scalable distributed data processing for ML workloads, streaming across CPU/GPU clusters and integrating with PyTorch, TensorFlow, and Ray Train. Use it for batch inference, ETL, and multi-modal preprocessing from a laptop up to hundreds of nodes.

LiteParse is a fast, open-source, and local document parser designed for layout-aware text extraction from PDFs, images, and office files. It outputs structured JSON with bounding boxes and metadata, making it ideal for building advanced RAG systems, citation grounding, and complex data pipelines without needing cloud APIs. Features include OCR, batch processing, and page screenshot capture.

This skill automatically converts structured data from Excel (.xlsx) workbooks into a clean, readable Markdown format. It is essential for enabling accurate analysis, summarization, and data extraction of spreadsheet contents, especially when the user asks to 'read', 'analyze', or 'compare' data without specifying a conversion method. It supports both single files and batch processing of entire folders.

This skill converts Microsoft Word documents (.docx) into Markdown format. It is essential for accurately analyzing, summarizing, or extracting data from proprietary Word files, as direct parsing is unreliable. It supports single files, batch processing of folders, and is designed to coordinate with sibling skills (PDF and Excel) when handling mixed document types to ensure no data is skipped.