Scholar SearchSAFE
An MCP server for academic paper search that integrates with AI assistants (e.g., Claude Code, Cursor), enabling them to search and retrieve academic paper metadata.
Overview
From the repository's own README, as read at the audited commit. Badges and raw HTML are left out.
An MCP server for academic literature workflows in Claude, Cursor, and other MCP clients.
It combines Semantic Scholar and arXiv into one unified toolset, with fast parallel search, normalized outputs, source-aware deduplication, and practical research utilities (citations, references, author graph, recommendations, and arXiv source download).
Table of Contents
- Why this project
- Demo videos
- Install
- Quick setup (Claude Desktop / Cursor)
- Environment variables
- Tool list
- Testing with MCP Inspector
- Contributing
Why this project
Most paper tools force you to choose one source or one API style. scholar-search-mcp provides one MCP layer for literature search and graph retrieval:
- One MCP server, multiple scholarly sources
- Free-first defaults (
arXivworks without keys) - LLM-friendly outputs for downstream reasoning and agent workflows
- Practical research actions, not only search
- Unified search:
search_papersruns Semantic Scholar + arXiv in parallel and deduplicates by normalized title. - Research graph tools: details, citations, references, author profile/papers, and recommendations.
- Batch + source workflows: fetch up to 500 papers, and download/extract arXiv LaTeX sources.
- Operational controls: built-in caching plus env-based source toggles (enable/disable channels).
- Source strategy: built-in Semantic Scholar + arXiv, free-first by default (
arXivkey-free), optional API key for higher Semantic Scholar limits.
Demo videos
Agent writes a survey paper with Scholar Search MCP.
##
6a67cadd4afaOBSERVED · 2026-10-06Connect
Built from this server's own package name, version and transport as found in its source — not copied from anyone's documentation, so it cannot drift against a page we do not control. Replace the environment placeholders with a token scoped to the least it needs.
claude mcp add scholar-search-mcp --env SEMANTIC_SCHOLAR_API_KEY=${SEMANTIC_SCHOLAR_API_KEY} -- uvx scholar-search-mcp{
"mcpServers": {
"scholar-search-mcp": {
"command": "uvx",
"args": [
"scholar-search-mcp"
],
"env": {
"SEMANTIC_SCHOLAR_API_KEY": "${SEMANTIC_SCHOLAR_API_KEY}"
}
}
}
}Exposed tools (8)
8 read · 0 write · 0 destructive.
| Tool | Risk | Description |
|---|---|---|
batch_get_papers | read | Get details for multiple papers (up to 500). |
get_author_info | read | Get author details. |
get_author_papers | read | Get papers by author. |
get_paper_citations | read | Get list of papers that cite this paper. |
get_paper_details | read | Get paper details. Supports DOI, ArXiv ID, Semantic Scholar ID, or URL. |
get_paper_recommendations | read | Get similar paper recommendations for a paper. |
get_paper_references | read | Get list of references of this paper. |
search_papers | read | Search academic papers by keyword. Optional filters: year, venue. |
Trust audit
SAFEgrade B · trust 89/100 Nothing in the source contradicts what it says it does. Grade A is reserved for packages that have also passed the behavioural sandbox.
| Layer | What it checks | Result |
|---|---|---|
| L0 | Provenance & inventory | PASS |
| L1 | Static analysis of the code | PASS |
| L2 | Instruction surface (what it tells the agent) | PASS |
| L3 | Class-specific surface | PASS |
| L4 | Behavioural (sandbox) | SKIPPED |
What the source does
- Filesystem
- declared (1 observation(s))
- Network
- declared (4 observation(s))
- Shell
- none-observed
- Dependencies
- not all pinned
- Secrets in source
- none-found
Findings (1)
mcp, httpx, diskcache
Gates applied: no_behavioural_pass.
6a67cadd4afafull audit observations/trust-audit/mcp-server/silung__scholar-search.json · Report an issue / request a re-scanAudit history
Every audit this server has had. A grade with a past is a grade somebody is still checking.
| Date | Source | Verdict | Grade | Score | Change |
|---|---|---|---|---|---|
| 2026-10-06 | 6a67cadd4afa | SAFE | B | 89 | first audit |
Questions
What is the Scholar Search MCP server?
An MCP server for academic paper search that integrates with AI assistants (e.g., Claude Code, Cursor), enabling them to search and retrieve academic paper metadata.
What tools does Scholar Search expose?
8 in total: 8 read-only, 0 that write, and 0 that can delete or overwrite. Every one is listed on this page with its risk.
Is Scholar Search safe to connect to an agent?
The audit found nothing in the source that contradicts what it says it does, and graded it B (89/100). Grade A is held back for packages that have also passed a sandboxed behavioural run, which is why a clean server reads B.
What credentials does Scholar Search need?
It reads SEMANTIC_SCHOLAR_API_KEY from the environment. Give it a token scoped to the least it needs — an agent that can be talked into calling a tool can be talked into calling it with your credentials.
How does Scholar Search run?
It speaks stdio, so it runs as a local process your client starts. It is published on PyPI as scholar-search-mcp.
How current is this page?
The grade is for one exact copy of the source (6a67cadd4afa), read on 2026-10-06. The repository is watched and re-audited when it changes.