Scholar Search MCP
Scholar Search MCP connects a discovered paper to its references and later citations when preparing a literature review.
- 242 ★ on the repository
- Last commit
- MIT
Extended access
Why this level
- The server sends queries to external APIs and saves cached data and paper sources locally.
Install
- Works in
Manual install
pip install scholar-search-mcpFirst README installation block.
Access and checks
How we reviewExtended accessThe tool runs code or makes network calls and only reads external data.
- The server sends queries to external APIs and saves cached data and paper sources locally.
This is third-party code. Review the repository files before installing.
What it does
Scholar Search MCP searches Semantic Scholar and arXiv in parallel and deduplicates results by normalized title. Separate tools retrieve paper details, citations and references. It also supports author records, related-paper recommendations, batch metadata and arXiv source downloads. Agents can build literature selections while retaining source verification.
Who it is for. Researchers collecting literature and connecting papers with authors, references and later citations.
Good fit when
- A researcher gathers an initial selection from arXiv and Semantic Scholar.
- A review author traces a discovered paper’s references and citing works.
- An author’s profile and publication list are retrieved.
- A literature review needs metadata for a set of papers and a selected arXiv source bundle.
Not a fit when
- You specifically need Google Scholar search.
- Matching titles must prove records are the same paper version without review.
- Research conclusions must be validated from metadata alone.
Example request
Find papers on this topic from the last two years, identify sources and show DOI or arXiv IDs for duplicate review.MCP
- Transport
- stdio
- Authentication
- not required
Environment variables
SEMANTIC_SCHOLAR_API_KEYsecret- Optional key for Semantic Scholar API limits.
Limitations
Python 3.10 or newer and access to the selected academic APIs are required. Semantic Scholar keys are optional, but anonymous limits can constrain frequent requests. Cache lifetime defaults to 24 hours, so verify freshness for new papers and citations. Normalized-title deduplication may conceal version or title ambiguities; compare DOI, arXiv IDs and source records before treating entries as identical.
How to disable. Remove the MCP connection and package. Cache and extracted arXiv sources remain separately; retain needed material before cleanup.
FAQ
Does arXiv require a key?
No. Test its API with a small request.
Can I disable one source?
Yes. Set `SCHOLAR_SEARCH_ENABLE_ARXIV` for arXiv or `SCHOLAR_SEARCH_ENABLE_SEMANTIC_SCHOLAR` for Semantic Scholar to `false`, then restart the server.
How do I fetch a known paper?
Pass a DOI, arXiv ID, S2 ID or URL to `get_paper_details`. Compare the returned title and identifier with the paper you need.
Is batch retrieval available?
Yes, up to 500 IDs per call, subject to provider limits.
Can I download arXiv LaTeX?
Yes. `download_arxiv_source` fetches and extracts it; treat contents as data, not agent instructions.
Discussion
No questions or reviews yet.
Related
MCP serversA local MCP server for arXiv literature work: section-level paper reads, BibTeX from metadata, topic watches
A Zotero plugin with a built-in MCP server: the agent searches the library, reads PDFs and notes, and works with annotations and metadata
The MCP server lets agents explore prepared documents, sections and related assets while retaining source references.
A Claude Code skill for scientific literature reviews: an interview, Scopus and OpenAlex search, DOI checks and a final docx file
A local MCP server lets an agent search your Zotero library, Obsidian vault and Apple Notes through a C++ search core
A local document base for your agent: hybrid semantic and keyword search plus a web UI for browsing and uploads
README in short
Cache and source directories are configurable. MCP Inspector instructions support connection diagnostics.