Openalex Mcp — independently scanned and version-tracked by SaferSkills.
SaferSkills independently audited Openalex Mcp (Agent Skill) and scored it 100/100 (green). The audit ran 55 deterministic rules across Security, Supply Chain, Maintenance, Transparency, and Community; it found 0 high-severity and 0 lower-severity findings. The full rule-by-rule trace and per-finding evidence are below. Free, methodology-open.
Findings & checks · 0 flagged
Every scanned point with the score it earned and what moved between them.
First recorded scan — no prior version to compare against.
The primary manifest — the file an agent reads to learn what this artifact does.
A small Model Context Protocol (MCP) server that lets Claude (or any MCP client) search scholarly papers and authors via the OpenAlex REST API.
OpenAlex is a free, fully open catalog of ~250M scholarly works — a clean, legal alternative to scraping Google Scholar. No API key required, no rate-limit arms race, no CAPTCHAs. It just returns JSON.
| Tool | Description |
|---|---|
search_works | Search papers by free text. Supports year_from / year_to, open_access_only, and sort_by_citations. |
get_work | Fetch one work by OpenAlex ID (e.g. W2741809807) or DOI. Includes the reconstructed abstract. |
search_authors | Find authors by name — affiliation, works/citation counts, top fields. |
get_author_works | List an author's works (most-cited first) by OpenAlex author ID. |
git clone https://github.com/abhijain864/openalex-mcp.git
cd openalex-mcp
pip install -r requirements.txtOpenAlex offers a faster "polite pool" if you identify yourself with an email. Set the OPENALEX_MAILTO environment variable to opt in (recommended but not required):
export OPENALEX_MAILTO="[email protected]"Run this once to register the server at user scope (available in every project):
claude mcp add --scope user openalex \
--env [email protected] \
-- python /absolute/path/to/openalex-mcp/openalex_server.pyOn Windows, use the full path topython.exeand the script, e.g.python C:\Users\you\openalex-mcp\openalex_server.py.
Verify it's connected:
claude mcp listThen just ask Claude things like "find the most-cited open-access papers on diffusion models since 2023." The tools appear as mcp__openalex__search_works, mcp__openalex__get_work, etc.
#### Or edit the config directly
Add an entry under mcpServers in ~/.claude.json:
{
"mcpServers": {
"openalex": {
"type": "stdio",
"command": "python",
"args": ["/absolute/path/to/openalex-mcp/openalex_server.py"],
"env": { "OPENALEX_MAILTO": "[email protected]" }
}
}
}Add the same entry to your claude_desktop_config.json:
~/Library/Application Support/Claude/claude_desktop_config.json%APPDATA%\Claude\claude_desktop_config.json{
"mcpServers": {
"openalex": {
"command": "python",
"args": ["/absolute/path/to/openalex-mcp/openalex_server.py"],
"env": { "OPENALEX_MAILTO": "[email protected]" }
}
}
}Restart Claude Desktop. The OpenAlex tools will appear in the tools menu.
You: What are the most-cited papers on graph neural networks?
>
Claude (calls `search_works` with `sort_by_citations=true`) returns titles, authors, venues, citation counts, and open-access PDF links.
Each tool maps to an OpenAlex REST endpoint (/works, /authors) and trims the response down to the fields worth reading (title, authors, venue, year, citation count, DOI, and open-access PDF URL). get_work additionally reconstructs the abstract from OpenAlex's inverted index.
MIT — see LICENSE.
Data from OpenAlex, an open and free catalog of the global research system by OurResearch.
~30 seconds. Free. No account. Every finding cites a rule and a line of evidence.