Does it require an existing list of papers?

Yes, it primarily uses a papers/core_set.csv file as input to ensure categories are based on actual evidence rather than abstract concepts.

How does it prevent generic categories?

The skill employs Scope Guardian logic to avoid 'Misc' or 'Other' buckets, requiring specific inclusion cues and representative paper IDs for every node.

What is the main output of the Taxonomy Builder skill?

It generates a structured outline/taxonomy.yml file containing a 2+ level hierarchical structure with specific descriptions and inclusion cues for each node.

Can I use this for non-academic technical writing?

Yes, it is ideal for any technical snapshot, product comparison, or documentation project that requires a deep, structured classification of information based on a set of source documents.

What happens if the generated taxonomy is too shallow?

The skill includes a built-in quality gate that flags taxonomies with fewer than two levels or those containing 'TODO' and placeholder text, preventing low-quality outputs.

Taxonomy Builder

Name: Taxonomy Builder
Author: WILLOSCAR

byWILLOSCAR

•

데이터 과학 및 ML

Generates structured, multi-level research taxonomies from paper sets to anchor survey outlines and content mapping.

Taxonomy Builder is a specialized utility for researchers and technical writers that transforms a collection of core papers into a hierarchical theme structure. By leveraging an evidence-first methodology, it ensures that every category is grounded in actual data rather than generic placeholders. It facilitates the creation of 2+ level taxonomies that serve as the backbone for literature reviews, snapshots, and state-of-the-art reports, ensuring each node is distinct, mappable, and reader-friendly through role-based guidance from Architect, Sponsor, and Guardian perspectives.

주요 기능

01Integrated quality gates to flag placeholders and ensure structural integrity

02Generates standardized taxonomy.yml output with detailed node descriptions

03Automated hierarchy generation (2+ levels) from paper metadata

04Evidence-first categorization to prevent generic or hollow buckets

05Role-based logic for architectural depth and scope guarding

0688 GitHub stars

사용 사례

01Organizing literature reviews for academic or technical surveys

02Creating structured outlines for domain-specific state-of-the-art reports

03Mapping research papers to specific design and evaluation axes

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add willoscar/research-units-pipeline-skills taxonomy-builder

For use in Claude.ai and ChatGPT

Download Skill