Four-Dimensional Deep Reading

Problem

Reading a book often fails at the reuse stage: the main claims, methods, failure conditions, counterexamples, and comparative context stay mixed in memory. four-dimensional-deep-reading turns a book title, file, or link into a structured deep-reading report for engineers and researchers, reducing ad hoc summaries and fragmented notes.

How It Works

It first enriches the input with background data, then runs four roles in parallel:
- Axiom Analyst: reduces viewpoints to atomic propositions and marks falsifiable assumptions such as A1 and A2.
- Structured Note Taker: outputs chapter tables, relationship networks, key turning points, and compresses the result into Logic, Method, and Summary.
- Professional Contrarian: looks for black-swan cases, boundary conditions, and failure modes so conclusions are not accepted uncritically.
- Random Role: samples one identity from a preset pool to add a cross-domain angle missed by the first three roles.

The report usually includes book-review search, publication metadata, chapter structure, horizontal comparison, and diachronic timeline. With only a title, the skill routes by language to sources such as Douban, Goodreads, Amazon, Booklog, and Yes24. With a local file, it attempts to parse txt, pdf, epub, mobi, and md, then cleans chapter markers.

Boundaries

It fits long-form books, documents, or chaptered materials where deep decomposition is useful. It is not ideal for pure code review, live fact-checking, or scholarly writing that requires verbatim citations. External reviews and metadata depend on platform accessibility, so treat the report as an analytical draft rather than a definitive source.