yifanfeng97/hyper-extract — reverse-engineered prompt

Build me a Python command line tool that can take messy documents like markdown, PDFs, and notes, then turn them into structured knowledge I can actually explore. I want one simple flow where I can parse a document into things like lists, typed records, knowledge graphs, hypergraphs, and spatio temporal graphs, depending on the template I choose.

It should feel useful right away for real cases like research papers, company reports, and general biography style documents. After parsing, I want to search the extracted knowledge in plain English, open a simple visualization of the graph, and export the result as an Obsidian vault with linked markdown notes. Please support growing the knowledge base over time by adding more documents later instead of starting over.

Make the CLI friendly and polished, with a config step for API keys, sensible defaults, and a small Python API too. If possible, include a bunch of ready to use YAML templates across common domains. Look up current docs online if you need to.