Live OpenAlex access for MCP agents and literature analysis
openalex-mcp-server, developed by Cyanheads, is an MCP server that connects AI agents to the OpenAlex academic catalog for live research grounding and citation analysis. The server lets models query and retrieve metadata, run citation-graph analysis, and generate structured literature-review prompts against OpenAlex data. Key capabilities include abstract reconstruction, entity resolution, and realtime budget reporting. It targets academic researchers, data scientists, and AI developers who need agent-native access to scholarly metadata.
What researchers can actually do with the server
The server gives AI assistants programmatic access to the OpenAlex index, enabling searches across more than 270 million works, authors, and institutions. Agents can perform targeted scholarly queries using the full OpenAlex grammar and follow built-in literature-review prompts to synthesize findings. Practical outputs include:
- focused literature reviews guided by agent prompts
- citation-graph mapping and trend analysis
- resolving ambiguous names to DOIs or OpenAlex IDs
How reliable the generated research material is
The server reconstructs abstracts by converting position-keyed inverted indices into plaintext and applies data normalization for DOIs, ORCIDs, RORs, and PMIDs, which improves record linking. Those parsing and normalization steps reduce mismatched identifiers, but agent-produced syntheses still require checking against the original OpenAlex records for factual assertions and citation accuracy.
How it fits into existing agent workflows and setup needs
The server runs on any host that supports the Model Context Protocol, listing compatibility with Claude Desktop, Cursor, and Smithery, and it accepts STDIO or streamable HTTP connections. Runtime requirements state Node.js v24+ or Bun v1.3.0+. It reports API usage against the OpenAlex daily budget and supports an optional API key or email for the polite pool, helping manage rate limits in agent-driven workflows.
Suitable for technically proficient teams that need live scholarly data
openalex-mcp-server is a practical option for research groups and AI developers who integrate agent workflows with live bibliographic data; its design aligns with MCP-based deployments and the mcp-ts-core TypeScript framework. Expect to operate and tune the server within an existing developer environment, and pair agent syntheses with manual verification of primary OpenAlex records for high-stakes claims.
