crossref-mcp
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@crossref-mcpfind the paper with DOI 10.1038/nature12373"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
crossref-mcp
Crossref REST API を使って学術文献のメタデータを取得する MCP サーバーです。
機能 (Tools)
Tool | 説明 |
| DOI から1件の文献メタデータを取得します。 |
| フリーテキスト / タイトル / 著者から文献を検索します。 |
取得・検索結果には、タイトル・著者・掲載誌・出版年・DOI・被引用数・(提供されていれば)アブストラクトなどが含まれます。
Related MCP server: CrossRef MCP Server
実行 (uvx)
公開リポジトリから直接実行できます:
uvx --from git+https://github.com/ebiyu/crossref-mcp crossref-mcpローカルのクローンから実行する場合:
uvx --from . crossref-mcpMCP クライアントへの登録
Claude Desktop / Claude Code などの mcpServers 設定例:
{
"mcpServers": {
"crossref": {
"command": "uvx",
"args": ["--from", "git+https://github.com/ebiyu/crossref-mcp", "crossref-mcp"],
"env": {
"CROSSREF_MAILTO": "you@example.com"
}
}
}
}ローカルにクローンしたものを使う場合は、--from に絶対パスを指定します(Windows の例):
{
"mcpServers": {
"crossref": {
"command": "uvx",
"args": [
"--from",
"PATH_TO_THIS_FOLDER",
"crossref-mcp"
],
"env": {
"CROSSREF_MAILTO": "you@example.com"
}
}
}
}macOS / Linux の場合は "/path/to/crossref-mcp" のように指定してください。
Claude Code の CLI から登録する場合:
claude mcp add crossref -e CROSSREF_MAILTO=you@example.com -- uvx --from git+https://github.com/ebiyu/crossref-mcp crossref-mcpレート制限の尊重
Crossref の アクセスポリシー に従い、自動でリクエスト間隔を調整します。
レスポンスの
X-Rate-Limit-Limit/X-Rate-Limit-Intervalヘッダから許容レートを学習し、間隔 ÷ 上限秒ずつリクエストを空けます。リクエストは直列化され、同時実行数の上限も超えません。429 Too Many Requestsを受けた場合はRetry-Afterを尊重し、なければ指数バックオフ(最大30秒、最大5回)で再試行します。
CROSSREF_MAILTO(推奨)
環境変数 CROSSREF_MAILTO にメールアドレスを設定すると、Crossref の
"polite pool"
が使われ、より安定したレスポンスが得られます。設定は任意です。
開発
uv sync
uv run crossref-mcp # stdio でサーバー起動ライセンス
MIT
Available Tools
2 toolsget_work_by_doiA
Retrieve bibliographic metadata for a single work by its DOI.
Args: doi: The Digital Object Identifier, e.g. "10.1145/3292500.3330701". A full URL such as "https://doi.org/10.1145/..." is also accepted.
Returns: A dictionary of the work's metadata (title, authors, journal, year, DOI, abstract when available, citation counts, etc.).
| Name | Required | Description | Default |
|---|---|---|---|
| doi | Yes |
Output Schema
| Name | Required | Description |
|---|---|---|
No output parameters | ||
search_worksA
Search Crossref for works by free text, title, and/or author.
At least one of query, title, or author must be provided.
Args: query: Free-text bibliographic search across all fields. title: Restrict the search to the work's title. author: Restrict the search to author names. rows: Number of results to return (1-50, default 5).
Returns:
A dictionary with total_results and a works list of metadata
dictionaries, each in the same shape as get_work_by_doi.
| Name | Required | Description | Default |
|---|---|---|---|
| rows | No | ||
| query | No | ||
| title | No | ||
| author | No |
Output Schema
| Name | Required | Description |
|---|---|---|
No output parameters | ||
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
2 tool updates
v0.1.0- First observed
get_work_by_doi - First observed
search_works
TDQS
Scored across 2 tools
The two tools are clearly distinct: one retrieves a specific work by DOI, the other searches for works using query parameters. There is no overlap in their purposes, making it easy for an agent to select the appropriate tool.
Both tools follow a consistent verb_noun pattern: 'get_work_by_doi' and 'search_works'. The naming clearly indicates the action and the resource, and there is no mixing of conventions.
With only two tools, the server feels thin for a general Crossref API, but for the narrow purpose of retrieving works by identifier or search, it is minimally sufficient. The count is at the low end of what is reasonable.
The core workflows of looking up a work by DOI and searching for works are covered. Minor gaps exist, such as no support for other identifiers or listing all works of an author directly, but the primary use cases are addressed.
Maintenance
Related MCP Connectors
Resolve DOIs, search ~188M scholarly works, and fetch references via the Crossref REST API.
Search 150M+ academic works, journals, and funders via Crossref API.
Crossref MCP — wraps the Crossref REST API (academic papers, free, no auth)
Verify a DOI or find papers via Crossref metadata. $0.005/call via x402.
Related MCP Servers
- FlicenseNot gradedqualityDmaintenanceEnables AI assistants to search academic literature and retrieve detailed metadata for papers, journals, and funding organizations via the Crossref database. It supports DOI-based lookups and keyword searches to facilitate programmatic access to scholarly research data.7-
- AlicenseNot gradedqualityDmaintenanceSearches and retrieves scholarly metadata from the CrossRef REST API, covering over 150 million records across all disciplines, without requiring an API key.MIT
- FlicenseAqualityDmaintenanceEnables searching and retrieving scholarly metadata from Crossref's public REST API, including works, journals, funders, and members, without requiring an API key.101-
- AlicenseAqualityCmaintenanceEnables scholarly metadata lookups from the Crossref REST API, including works, members, journals, funders, types, licenses, and prefixes, as tools for LLM clients.1839 PyPIMIT