Search arXiv, fetch metadata, and read full paper content.
This MCP tool has a clear purpose: searching arXiv papers, fetching metadata, and reading full-text content. With an official registry listing, open-source code, and recent maintenance, no clear high-risk red flags are evident, but it should still be used with least-privilege controls because it runs locally and makes network requests.
The listed environment variables are a base URL, request delay, timeout settings, and log level; no API key, token, or other highly sensitive credential is indicated. Credential abuse risk appears low, though runtime configuration should still not be exposed in logs or shared configs.
The tool is declared to connect to arxiv.caseyjhand.com to search and retrieve paper content, which is consistent with its stated function and a known endpoint. User queries and requested paper content may be sent to that endpoint, but the materials do not indicate exfiltration to unrelated or unknown servers.
The system checks indicate that this MCP tool executes code or runs a local process, which is a normal characteristic of MCP tools. The provided materials do not show requests for system privileges beyond what would typically be needed for paper retrieval, nor do they reveal an obviously dangerous execution chain.
The description only indicates access to paper metadata and full-text content, with no explicit claim of broad local file access, user-directory access, or writes to sensitive resources. However, as a locally running MCP process, its filesystem and context visibility should still be constrained by default.
The source is an official registry entry and provides an open-source repository, giving it reasonable auditability. It has been updated within the last year, with no clear sign of abandonment. Although community stars are low and the license is not declared, that adds some adoption and compliance uncertainty but does not by itself amount to a high-risk red flag.
Copy the install command and let the AI configure it · recommended for beginners
Please install the "io.github.cyanheads/arxiv-mcp-server" MCP server from askskill: Run: claude mcp add 'io-github-cyanheads-arxiv-mcp-server' -- npx -y @cyanheads/arxiv-mcp-server
Search arXiv papers, fetch metadata, browse categories, and read full text.
Search arXiv papers, retrieve metadata, and access full PDFs.
Search, download, analyze arXiv papers, and save AI-generated summaries.
Search ArXiv papers, inspect details, track categories, and explore author networks.
Search arXiv papers and retrieve full content in Markdown for analysis.
Search, read, and extract key insights from arXiv papers via AI assistants.