mcp-scholar
# MCP Scholar
基于MCP协议的谷歌学术搜索和分析服务。
## 功能特点
- 谷歌学术论文搜索：根据关键词搜索相关论文，并按引用量排序
- 学者主页分析：分析谷歌学术个人主页，提取引用量最高的论文
- 支持与所有支持MCP客户端集成
- 支持与Cherry Studio集成：可以作为插件在Cherry Studio中使用
## 安装方法
### 启动服务器
```bash
# 方式一：使用uvx启动
uvx mcp-scholar
# 方式二：clone仓库后使用uv run启动
uv --directory 路径\到\mcp_scholar run mcp-scholar
```
### 在Cherry Studio中使用
- 「参照官方教程：https://vaayne.com/posts/2025/mcp-guide 」
## 示例用法
在Cherry Studio中，可以使用以下提示：
- 「总结5篇关于人工智能的论文」
- 「分析学者主页 https://scholar.google.com/citations?user=xxxxxx 的前10篇高引论文」
## 开发说明
本项目使用MCP协议开发，基于Python SDK实现。详细信息请参考[MCP Python SDK](https://github.com/modelcontextprotocol/python-sdk)。
## 许可证
MITTDQS
Scored across 7 tools
There is significant overlap between adaptive_search and scholar_search, both performing Google Scholar searches with similar parameters, which creates ambiguity. Additionally, summarize_papers appears to combine searching and summarization, potentially overlapping with the search tools, though its summarization aspect provides some distinction.
Most tools follow a clear noun_verb or verb_noun pattern (e.g., adaptive_search, paper_detail, summarize_papers), with consistent snake_case throughout. The only minor deviation is health_check, which fits the pattern but stands out as a utility tool rather than a core scholarly function.
With 7 tools, the count is reasonable for a scholarly search server, covering key operations like searching, retrieving details, and summarizing. However, the overlap between some tools suggests the set could be streamlined without losing functionality, making it slightly over-scoped.
The toolset covers core scholarly functions well, including searching, retrieving paper details, references, and scholar profiles, with summarization as an advanced feature. A minor gap is the lack of tools for managing saved papers or citations, but the existing tools support most common workflows without dead ends.