Local LLM API Cookbook: Build Scraping Agents That Work Offline

This cookbook teaches you how to build intelligent scraping agents using your local LLMs (Ollama/MLX/llama.cpp) by integrating open-source tools like Scrapy, Trafilatura, and Crawlee — no internet required. You'll learn how to connect APIs, extract structured data, and automate workflows using a curated registry of compatible tools. Before building agents, ensure your local LLM (Ollama/MLX/llama.cpp) can handle natural language prompts and JSON outputs. Test with a simple prompt like: > "Extract the title and URL from this HTML block." Select tools based on use case: - **Scrapy**: For robust, scalable web scraping.