OpenAIRE MCP Server

OpenAIRE MCP Server

Enables AI assistants to explore the OpenAIRE Research Graph and ScholExplorer citation index, discovering citation relationships and retrieving metadata for publications, datasets, and software.

Category
访问服务器

README

OpenAIRE MCP Server

<!-- SPDX-License-Identifier: Apache-2.0 -->

MCP (Model Context Protocol) server for exploring the OpenAIRE Research Graph and ScholExplorer citation index. Enables AI assistants like Claude to discover citation relationships, find related research outputs, and retrieve comprehensive metadata for publications, datasets, and software.

Table of Contents

Overview

This MCP server provides AI assistants with access to two powerful research discovery APIs:

  • ScholExplorer: Citation index with 3.6 billion scholarly relationships linking publications, datasets, and software
  • OpenAIRE Research Graph: Comprehensive metadata aggregator covering millions of research outputs with full abstracts, author affiliations, citation metrics, and funding information

Features

  • 🔗 Find all research outputs citing a specific DOI (publications, datasets, software)
  • 📊 Retrieve comprehensive metadata from OpenAIRE Research Graph
  • 🤖 Seamless integration with Claude Desktop and other MCP clients
  • 🐳 Docker containerization support
  • 📝 Structured JSON logging
  • ⚙️ Configuration via environment variables (12-factor app)

Available Tools

  1. find_related_research - Find publications, datasets, and software citing a given DOI
  2. get_metadata - Get detailed metadata for research products from OpenAIRE Graph

Installation

Requirements

  • Python 3.10 or higher
  • pip

Install from Source

# Clone the repository
git clone https://github.com/cessda/cessda.ai.mcp.openaire.git
cd cessda.ai.mcp.openaire

# Install in development mode
pip install -e .

Verify Installation

mcp-openaire --help

Configuration

The server follows the 12-factor app methodology and is configured via environment variables.

Environment Variables

Variable Default Description
OPENAIRE_SCHOLEXPLORER_URL https://api.scholexplorer.openaire.eu/v3 ScholExplorer API base URL
OPENAIRE_GRAPH_URL https://api.openaire.eu/graph/v2 OpenAIRE Graph API base URL
OPENAIRE_API_TIMEOUT 30 API request timeout (seconds)
OPENAIRE_API_MAX_RETRIES 3 Maximum retry attempts
OPENAIRE_LOG_LEVEL WARN Logging level (INFO, WARN, ERROR)
OPENAIRE_DEFAULT_LIMIT 200 Default citation result limit
OPENAIRE_MAX_LIMIT 1000 Maximum citation result limit
OPENAIRE_PAGE_SIZE 50 Results per API page request

Configuration File (Optional)

Create a .env file in your working directory:

OPENAIRE_API_TIMEOUT=60
OPENAIRE_LOG_LEVEL=INFO
OPENAIRE_DEFAULT_LIMIT=100

Usage

Claude Desktop Setup

  1. Locate your Claude Desktop configuration file:

    • macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
    • Windows: %APPDATA%\Claude\claude_desktop_config.json
    • Linux: ~/.config/Claude/claude_desktop_config.json
  2. Add the MCP server configuration:

{
  "mcpServers": {
    "openaire": {
      "command": "python",
      "args": ["-m", "mcp_openaire.server"],
      "env": {
        "OPENAIRE_LOG_LEVEL": "INFO"
      }
    }
  }
}

Or if installed via pip:

{
  "mcpServers": {
    "openaire": {
      "command": "mcp-openaire"
    }
  }
}
  1. Restart Claude Desktop

  2. Verify the server is connected - Look for the MCP icon or check available tools

Available Tools

1. find_related_research

Find publications, datasets, and software that cite a given DOI.

Parameters:

  • doi (required): DOI to find citations for (e.g., "10.21338/NSD-ESS8-2016")
  • limit (optional): Max results (1-1000, default: 200)

Returns: Citation links from ScholExplorer with:

  • Total citations found
  • Array of citing objects with titles, authors, DOIs, publication dates
  • Relationship types (cites, references)
  • Object types (publication, dataset, software)
  • Link providers (OpenCitations, Crossref, DataCite)

2. get_metadata

Get comprehensive metadata for a research product from OpenAIRE Research Graph.

Parameters:

  • identifier (required): DOI or OpenAIRE ID (e.g., "10.2139/ssrn.3991520")

Returns: Enriched metadata including:

  • Full title and abstract
  • Complete author list with affiliations and ORCID IDs
  • Publication date and publisher
  • Citation metrics
  • Access rights and open access status
  • Funding information
  • Subject classifications
  • Access URLs

Example Prompts

Try these prompts in Claude Desktop:

  1. Find citations for a dataset:

    "Find all publications citing the European Social Survey dataset 10.21338/NSD-ESS8-2016"

  2. Discover related research:

    "Show me the first 50 citations for DOI 10.2139/ssrn.3991520"

  3. Get detailed metadata:

    "Get me the full metadata for DOI 10.21338/NSD-ESS8-2016 including abstract and authors"

  4. Explore citation network:

    "Find all research citing dataset 10.5281/zenodo.1234567 and show me the top 10 by date"

  5. Combined workflow:

    "Find datasets in CESSDA about climate change, then for the first result, find all publications that cite it and get metadata for the most recent one"

Development

Setup Development Environment

# Clone repository
git clone https://github.com/cessda/cessda.ai.mcp.openaire.git
cd cessda.ai.mcp.openaire

# Install with development dependencies
pip install -e ".[dev]"

Project Structure

mcp-openaire/
├── LICENSE.txt                    # Apache 2.0 license
├── README.md                      # This file
├── pyproject.toml                 # Python package configuration
├── Dockerfile                     # Container image (STDIO)
├── Dockerfile.http                # Container image (HTTP/SSE)
├── docker-compose.yml             # Docker Compose configuration
└── src/
    └── mcp_openaire/
        ├── __init__.py           # Package initialization
        ├── server.py             # FastMCP STDIO server
        ├── server_http.py        # FastMCP HTTP/SSE server
        ├── tools.py              # API integration logic
        ├── config.py             # Environment configuration
        └── logging_config.py     # Structured JSON logging

Code Standards

  • Style: PEP 8 (enforced via black and ruff)
  • Type Hints: Required for all public functions
  • Documentation: Docstrings for all modules, classes, and functions
  • Logging: Structured JSON to stdout
  • Configuration: Environment variables only (no hardcoded values)
  • Versioning: Semantic versioning (MAJOR.MINOR.PATCH)

Testing

# Run API integration tests
python test_api.py

# Format code
black src/

# Lint code
ruff check src/

Operations

Running in Docker

# Build image (STDIO mode)
docker build -t mcp-openaire:0.1.0 .

# Run container (STDIO mode)
docker run -i mcp-openaire:0.1.0

# Build HTTP/SSE image
docker build -f Dockerfile.http -t mcp-openaire-http:0.1.0 .

# Run HTTP server
docker run -d -p 8001:8000 mcp-openaire-http:0.1.0

# Using Docker Compose
docker compose up -d
docker compose logs -f
docker compose down

Logging

The server outputs structured JSON logs to stdout:

{
  "timestamp": "2025-01-15T10:30:45.123456Z",
  "level": "INFO",
  "message": "ScholExplorer search completed",
  "logger": "mcp_openaire",
  "doi": "10.21338/NSD-ESS8-2016",
  "total_available": 55,
  "results_returned": 55
}

Log Levels:

  • INFO: Informational messages (API requests, results)
  • WARN: Warnings requiring attention (validation issues, limits exceeded)
  • ERROR: Errors requiring investigation (API failures, exceptions)

Monitoring

Monitor the server using:

  • JSON log aggregation (ELK, Splunk, etc.)
  • Check for ERROR and WARN level messages
  • Track API response times and failure rates
  • Monitor citation search patterns

Troubleshooting

Issue: Server not appearing in Claude Desktop

  • Verify configuration file path and JSON syntax
  • Check server installation: which mcp-openaire
  • Review Claude Desktop logs for errors
  • Restart Claude Desktop after configuration changes

Issue: API timeouts

  • Increase OPENAIRE_API_TIMEOUT environment variable
  • Check network connectivity to OpenAIRE APIs
  • Review structured logs for timeout patterns
  • Reduce OPENAIRE_DEFAULT_LIMIT for faster responses

Issue: No citations found

  • Verify the DOI is correct and exists in ScholExplorer
  • Not all DOIs have citations yet (especially new publications)
  • Check that the DOI format is correct (starts with "10.")
  • Try a well-known DOI like "10.21338/NSD-ESS8-2016" to test connectivity

Issue: Metadata not found

  • Some DOIs may not be in OpenAIRE Graph yet
  • Try searching ScholExplorer first to get an OpenAIRE ID
  • Check the identifier format (DOI should start with "10.")

Versioning

This project uses semantic versioning:

  • MAJOR: Breaking changes to API or configuration
  • MINOR: New features, backward compatible
  • PATCH: Bug fixes, backward compatible

Current version: 0.1.0

License

Copyright © 2025 CESSDA ERIC

Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with the License. You may obtain a copy of the License at

http://www.apache.org/licenses/LICENSE-2.0

Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License.

See LICENSE.txt for full license text.

Links

Contributing

Contributions are welcome! Please open an issue or pull request.

推荐服务器

Baidu Map

Baidu Map

百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。

官方
精选
JavaScript
Playwright MCP Server

Playwright MCP Server

一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。

官方
精选
TypeScript
Magic Component Platform (MCP)

Magic Component Platform (MCP)

一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。

官方
精选
本地
TypeScript
Audiense Insights MCP Server

Audiense Insights MCP Server

通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。

官方
精选
本地
TypeScript
VeyraX

VeyraX

一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。

官方
精选
本地
graphlit-mcp-server

graphlit-mcp-server

模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。

官方
精选
TypeScript
Kagi MCP Server

Kagi MCP Server

一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。

官方
精选
Python
e2b-mcp-server

e2b-mcp-server

使用 MCP 通过 e2b 运行代码。

官方
精选
Neon MCP Server

Neon MCP Server

用于与 Neon 管理 API 和数据库交互的 MCP 服务器

官方
精选
Exa MCP Server

Exa MCP Server

模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。

官方
精选