OpenAlex MCP Server
An MCP server for OpenAlex: 31 tools for papers, citations, authors and journals, plus a skill with a small CLI for Claude Code.
- 60 ★ on the repository
- Last commit
- MIT
Minimal access
Why this level
- Queries go to the public OpenAlex API together with your email address if you set one.
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add mcp/openalex-mcp-serverDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewMinimal accessThe tool runs nothing and makes no network calls.
- Queries go to the public OpenAlex API together with your email address if you set one.
This is third-party code. Review the repository files before installing.
What it does
OpenAlex MCP Server gives an agent access to the open OpenAlex catalog of scholarly works. The `search_works`, `get_work` and `get_related_works` tools find papers and similar ones, `get_work_citations` and `get_work_references` follow citations forward and back, and `search_by_topic` explores a field. It has a TTL cache, retries with exponential backoff and input validation. Separate tools filter results by journal lists: presets such as UTD24, FT50, AJG and top AI conferences, plus institution groups. The same access ships as a skill in `skill/`: its `openalex` CLI costs no context until the skill is invoked.
Who it is for. Researchers and graduate students preparing a literature review who want an agent to search top journals and follow citations.
Good fit when
- You need a literature review limited to the FT50 or UTD24 journal lists.
- You must trace a key paper citations forward and backward.
- You need to choose a journal to publish in: h-index ranking and indexing status.
- In Claude Code a light skill is enough instead of an always-on server.
Not a fit when
- You need full paper text: OpenAlex returns metadata and abstracts.
- You must work offline: the server calls the OpenAlex API.
- You need paid databases such as Scopus: the catalog is open and its coverage differs.
Example request
Find recent supply chain work in FT50 journals and show who cites the most influential one.MCP
- Transport
- stdio
- Authentication
- not required
Environment variables
OPENALEX_EMAIL- Email address for the faster OpenAlex request pool.
OPENALEX_API_KEYsecret- Optional OpenAlex API key.
Limitations
The server reads the open OpenAlex API, so coverage depends on that catalog and there are no full paper texts. An email in `OPENALEX_EMAIL` is optional and moves requests to a faster pool, and `OPENALEX_API_KEY` is optional too. The `npx openalex-research-mcp setup` wizard writes the Claude Desktop config and checks connectivity, and the `--print` flag prints JSON for other clients. The skill in `skill/` suits agents with shell access such as Claude Code, while the server fits MCP clients better.
How to disable. Remove the `openalex-research-mcp` entry from the Claude Desktop config and restart the client. For the skill, take `skill/bin` out of `PATH` and unset `OPENALEX_EMAIL`.
FAQ
How do I connect the server to Claude Desktop?
Run `npx openalex-research-mcp setup`: the wizard finds the config, asks for an email and an optional key, writes the settings and checks connectivity. Then restart Claude Desktop.
What is the difference between the server and the skill?
The skill in `skill/` runs the small `openalex` CLI and takes no context until needed. The server is always connected and suits MCP clients.
Which journal presets exist?
UTD24, FT50, AJG and ABS tiers, top AI conferences and institution groups such as the Ivy League. The `list_journal_presets` tool lists them.
Is an OpenAlex key required?
No. An email speeds up requests and a key is optional. Both values go through `OPENALEX_EMAIL` and `OPENALEX_API_KEY`.
How can I check that it works?
Call the health check tool from the client: it is meant for monitoring how the server is doing.
Discussion
No questions or reviews yet.
Related
MCP serversAn MCP server for searching and downloading academic papers from arXiv, PubMed, bioRxiv, Semantic Scholar and other open sources
Official MCP server from the OpenSearch project exposes a cluster to agents: indexes, mappings, DSL search and shard state.
An MCP server and skill for Sciverse paper search: metadata, semantic retrieval, citations and full text for agents.
Elasticsearch MCP connects multiple search clusters, with an explicit target name and access scoped to the supplied credentials.
PubMed MCP searches biomedical papers, retrieves available full text and formats citations with retraction and correction notices.
Connects a Zotero library to Claude and other agents: search papers, annotate, cite, and search by meaning
README in short
The server is written in TypeScript and published on npm with a setup command. A light skill on the same data sits beside it because the author finds always-on MCP servers heavier than skills for routine queries.