The skill refines rather than invents. Starting from your research goals, research content and key scientific questions, it turns methods, technical route and innovation points into academic prose, improves their logic and formats them to convention. Its principles put rigor ahead of ornate wording, require a closed loop from content to questions to methods to route to innovation, keep your own materials and frameworks first and forbid carrying specifics from past proposals into the new project. Drafts, code, algorithm frameworks and flowchart sketches are optional inputs.
The workflow has seven steps: read inputs, analyze the logic chain and project type, design or optimize the methods, build the technical route, draw route diagrams as Mermaid code or design suggestions, distill innovation points by type and run a closed-loop check against a validation menu and anti-pattern list. The route is detailed down to sub-studies by default, with a coarser version only if you ask. Style guides and pattern files in a knowledge base folder govern each of the three sections.