semanticscholar-skill

Search Semantic Scholar for papers, citations, authors, and BibTeX metadata via its API.

2|Updated May 22, 2026
One-click install
npx skills add https://github.com/CDUTAKL/research-workflow-kit --skill semanticscholar-skill-cdutakl
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: semanticscholar-skill
Source: https://github.com/CDUTAKL/research-workflow-kit/tree/main/skills/semanticscholar-skill
Command: npx skills add https://github.com/CDUTAKL/research-workflow-kit --skill semanticscholar-skill-cdutakl

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires requests, and includes references (resource) components.

What problem does it solve? Literature discovery for papers, theses, and surveys often relies on memory or scattered manual searches, producing unverified citations. This Skill grounds academic search in real Semantic Scholar API responses, returning structured, deduplicated paper metadata ready for review. ## Core Features & Use Cases - Paper Search & Lookup: Run relevance-ranked, boolean bulk, full-text snippet, and exact-title searches, plus direct lookup by DOI, arXiv, PMID, or CorpusId. - Citation Graph & Recommendations: Traverse citing and referenced papers with intent labels, and get single-seed or multi-seed similar-paper recommendations. - Author Tools & Export: Search authors, list their papers, and export results as Markdown tables, JSON, or BibTeX with source_status and next_action fields for literature matrices. - Use Case: While writing a related-work section, ask for papers on a topic since 2018; the Skill runs one Python script against the API, deduplicates results, and returns a ranked table with DOIs, citation counts, and follow-up actions for Zotero and Scite. ## Quick Start Use the semanticscholar-skill to find recent papers on gamma spectrum classification with machine learning and export the top results as BibTeX.

Frequently Asked Questions about semanticscholar-skill

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I search Semantic Scholar for papers in Python?

Import the s2.py helper and call search_relevance() for broad topics or search_bulk() with build_bool_query() for precise boolean queries. Results return as paper dictionaries with title, year, authors, venue, externalIds, and citation counts.

How to find papers citing a specific paper using Semantic Scholar?

Call get_citations() with a paper ID such as a DOI, arXiv, or CorpusId prefix. It returns citing papers with contextsWithIntent labels showing whether each citation is methodology, background, or result oriented.

Do I need an API key for the Semantic Scholar API?

No, the S2_API_KEY environment variable is optional. Anonymous requests work with stricter rate limits, while a key raises throughput; the helper enforces a 1.1 second gap and exponential backoff either way.

Can I export Semantic Scholar results to BibTeX?

Yes, export_bibtex() concatenates the citationStyles.bibtex field across papers when that field is requested. The output should be treated as candidate BibTeX and verified against DOI, publisher, or arXiv sources before final citation.

Why does Semantic Scholar rate limiting cause request failures?

The API returns 429 or 504 errors under heavy load. The helper automatically retries up to five times with exponential backoff from 2 to 60 seconds, so persistent failures usually mean you need an S2_API_KEY for higher limits.

When should I not rely on Semantic Scholar metadata alone?

Semantic Scholar is a discovery tool, not a final citation authority. Verify titles, authors, venues, and DOIs through publisher pages, DOI resolvers, arXiv, PubMed, Zotero, or Scite before citing important references.