Doc Assistant Mcp
What is Doc Assistant Mcp
doc-assistant-mcp is a personalized document assistant server based on MCP that allows users to query the content of documents located on specified local paths.
Use cases
Use cases include querying specific information from academic papers, extracting data from spreadsheets, and retrieving text from various document formats for analysis or reference.
How to use
To use doc-assistant-mcp, start the server by running the command with specified document paths and an embedding model. Interact with the server using a compatible MCP client to query document contents.
Key features
Key features include support for multiple document formats (PDF, TXT, CSV, Excel, Word, Markdown), semantic search using a vector database, compliance with MCP protocol standards, and provision of standard MCP service interfaces.
Where to use
doc-assistant-mcp can be used in various fields such as research, education, and any environment where document retrieval and information extraction from local files are required.
Clients Supporting MCP
The following are the main client software that supports the Model Context Protocol. Click the link to visit the official website for more information.
Overview
What is Doc Assistant Mcp
doc-assistant-mcp is a personalized document assistant server based on MCP that allows users to query the content of documents located on specified local paths.
Use cases
Use cases include querying specific information from academic papers, extracting data from spreadsheets, and retrieving text from various document formats for analysis or reference.
How to use
To use doc-assistant-mcp, start the server by running the command with specified document paths and an embedding model. Interact with the server using a compatible MCP client to query document contents.
Key features
Key features include support for multiple document formats (PDF, TXT, CSV, Excel, Word, Markdown), semantic search using a vector database, compliance with MCP protocol standards, and provision of standard MCP service interfaces.
Where to use
doc-assistant-mcp can be used in various fields such as research, education, and any environment where document retrieval and information extraction from local files are required.
Clients Supporting MCP
The following are the main client software that supports the Model Context Protocol. Click the link to visit the official website for more information.
Content
个性化文档助手 MCP 服务器
一个基于 MCP 的个性化文档助手服务器，可以查询指定电脑路径的文档内容。
功能特点
- 支持多种文档格式：PDF、TXT、CSV、Excel、Word、Markdown 等
- 使用向量数据库进行语义检索
- 符合 MCP 协议标准
- 提供标准 MCP 服务接口
依赖安装
# 安装基础依赖
pip install langchain langchain-community faiss-cpu pypdf transformers unstructured docx2txt pandas openpyxl mcp
## 使用方法
### 启动服务器
```bash
python test_mcp_server.py --paths /path/to/documents /another/path --embedding-model moka-ai/m3e-base
参数说明：
--paths: 文档路径列表，可以是文件或目录--embedding-model: 嵌入模型名称，默认为 moka-ai/m3e-base
与服务器交互
该服务器支持标准的 MCP 交互协议，可以通过 MCP 客户端与其交互。
服务器提供的功能：
- 资源列表（list_resources）：列出所有可用的文档资源
- 读取资源（read_resource）：读取指定文档的内容
- 工具列表（list_tools）：列出可用的工具
- 调用工具（call_tool）：调用查询文档内容的工具
示例：查询文档
可以使用 MCP 客户端调用 query_document 工具来查询文档内容：
from mcp.client import Client
async with Client() as client:
# 连接到文档助手服务器
await client.connect("document_assistant_mcp")
# 调用查询工具
result = await client.call_tool(
"query_document",
{"question": "我想了解关于某个主题的信息"}
)
# 打印结果
print(result[0].text)
注意事项
- 首次加载文档时会构建向量索引，可能需要一定时间
- 查询性能取决于文档数量和嵌入模型的性能
- 服务器使用标准输入输出作为传输方式，适合与其他 MCP 组件集成使用
客户端配置
{
"mcpServers": {
"document-assistant": {
"command": "python",
"args": [
"你的 python 文件地址 如./mcp_server.py",
"--paths",
"你的文档路径 如/Users/Downloads",
"--embedding-model",
"moka-ai/m3e-base"
],
"env": {}
}
}
}Dev Tools Supporting MCP
The following are the main code editors that support the Model Context Protocol. Click the link to visit the official website for more information.
