Thurlow Research avatar

Thurlow Research

Official

@thurlow-research · United States of America

0Followers
|
5Public Repos
|
7Published Skills

Empirical research into scalable human oversight of AI-generated code — Purdue doctoral research by Scott Thurlow.

Skills Distribution
DomainData Systems...Bibliographic Data.. (40%)Citation Graph Tra.. (25%)Scholarly Metadata.. (20%)Full-Text Extraction (15%)

Agent Skills by Thurlow Research

Showing 7 vetted skills indexed across 1 GitHub repositories.

Frequently Asked Questions About Thurlow Research

FAQPage Schema
What tasks can I accomplish with Thurlow Research's skills?

You can run systematic literature reviews end-to-end: search Semantic Scholar and arXiv, snowball citations forward and backward, backfill missing abstracts and metadata from OpenAlex, query and export a Zotero library, deduplicate records losslessly, and convert PDF attachments to TXT for full-text analysis.

Who are these skills designed for?

Academic researchers and doctoral students conducting systematic literature reviews, particularly those managing Zotero libraries. They were built for Purdue doctoral research on governance of AI-generated code, but apply to any SLR requiring citation chasing, deduplication, and full-text screening.

How do the skills work together in practice?

arXiv runs query-only searches and hands a create-items plan to the zotero skill, which writes to the library. Semantic Scholar and OpenAlex enrich records and traverse citations, zotero-merge-prep consolidates duplicates before merging, and zotero-pdf-to-text generates TXT attachments for downstream reading.

Do these skills require API keys or paid access?

OpenAlex requires no API key and caches results locally. Semantic Scholar's citation graph requires a key. The zotero skill uses the Zotero Web API with your library credentials. arXiv querying is open. No licensing fees are indicated in the manifest.

What are the prerequisites and dependencies?

A Zotero personal or group library (optionally synced to local ~/Zotero/storage), pdftotext for PDF conversion, and a Semantic Scholar API key for citation chasing. The archived zotero-slr-dedup skill must not be executed; use zotero-merge-prep for all duplicate resolution.