Primo MCP Server
MCP server for searching Ex Libris Primo library catalogues and subscribed databases, enabling search, record retrieval, autocomplete, citation generation, and export to BibTeX, RIS, or CSV.
README
primo-mcp-server
MCP server for Ex Libris Primo library discovery -- search university catalogues and subscribed databases (ProQuest, Elsevier, Crossref, Gale, Springer, IEEE, etc.) via the Model Context Protocol.
Features
- Search the full university catalogue and Primo Central Index (millions of records)
- Get record details including abstract, authors, identifiers, and availability
- Autocomplete search suggestions
- Generate citations in APA 7th, Harvard, Chicago, IEEE, and Vancouver styles
- Export to BibTeX, RIS, or CSV for import into reference managers
Installation
git clone https://github.com/geheharidas/primo-mcp-server.git
cd primo-mcp-server
pip install -e .
Register in Claude Code
Add to ~/.claude/settings.json:
{
"mcpServers": {
"primo": {
"command": "python",
"args": ["-m", "primo_mcp_server"]
}
}
}
Restart Claude Code. The tools will appear as mcp__primo__primo_search, etc.
Tools
| Tool | Description |
|---|---|
primo_search |
Search the library catalogue with filters (type, date, peer-reviewed) |
primo_get_record |
Get full details for a record by ID |
primo_suggest |
Autocomplete search suggestions |
primo_cite |
Generate formatted citations (APA7, Harvard, Chicago, IEEE, Vancouver) |
primo_export |
Export records as BibTeX, RIS, or CSV |
Usage Examples
From a Claude Code conversation:
- "Search the library for articles about machine learning in entrepreneurship published after 2020"
- "Get the full details for record cdi_crossref_primary_10_1234"
- "Generate APA7 citations for these records"
- "Export the search results as BibTeX"
Configuration
Defaults are set for UWA (University of Western Australia). Override via environment variables:
| Variable | Default | Description |
|---|---|---|
PRIMO_BASE_URL |
https://onesearch.library.uwa.edu.au/primaws/rest/pub |
Primo API base URL |
PRIMO_VID |
61UWA_INST:NDE_UWA |
Primo View ID |
PRIMO_INSTITUTION_NAME |
UWA |
Display name |
PRIMO_REQUEST_TIMEOUT |
30.0 |
HTTP timeout in seconds |
PRIMO_MAX_RESULTS_PER_REQUEST |
50 |
Maximum results per search |
PRIMO_DEFAULT_RESULTS |
10 |
Default results per search |
See .env.example for the full list.
Running Tests
pip install -e ".[dev]"
pytest tests/ -v
Licence
MIT
推荐服务器
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。
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。
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。
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。
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。
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。
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。
mcp-server-qdrant
这个仓库展示了如何为向量搜索引擎 Qdrant 创建一个 MCP (Managed Control Plane) 服务器的示例。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器