Newbscraping: AI-Assisted Webscraping for Newbs (github.com)

Newbscraping is a beginner-focused repository that walks non-programmers through end-to-end web scraping using Cursor AI’s assistant. The repo uses a simple questionnaire (START_HERE.md) to capture your scraping goals, asks you to drop sample HTML/PDF/CSV files into html_input/file_input, then has Cursor AI generate and run Python scraping scripts that save structured CSV output to csv_output. The project layout includes AI_Instructions.md (technical prompts for the assistant), sample_scripts for reference, and clear step-by-step instructions; supported browsers are Chrome/Edge/Firefox/Brave (Safari not supported). Requirements are minimal: Python 3 and the Cursor IDE with its Python extension (new users get a Pro trial), while the AI handles the code generation so you don’t need to program. For the AI/ML community, Newbscraping showcases how LLM-driven tooling can automate data-collection pipelines—lowering the barrier for researchers and product teams to create labelled datasets, monitor web content, or prototype ETL flows without deep engineering. Important technical implications: generated scripts are Python-based and rely on developer-provided sample pages (so data quality depends on representative input), and users must still consider legal and ethical scraping constraints, site terms, and anti-bot protections. Overall, it’s a practical demo of how AI can scaffold reproducible scraping workflows and accelerate dataset creation for ML experiments.