Alumnx MCP Server
Provides agricultural intelligence tools for pest/disease lookup, government scheme discovery, and SME knowledge retrieval via semantic search.
README
Alumnx MCP Server
A Model Context Protocol (MCP) server for agricultural intelligence, providing tools for pest/disease lookup, government scheme discovery, and SME knowledge retrieval via semantic search.
Features
- Pests & Diseases — RAG-powered lookup for crop pest and disease information
- Government Schemes — RAG-powered search for agriculture-related government schemes by type and state
- SME Divesh — Semantic search over a Pinecone knowledge base (SME-Divesh namespace)
- MCP + REST — Dual interface: MCP protocol (via FastMCP) and plain REST (
/callTool)
Requirements
- Python 3.9+
- Dependencies (install via
pip install -r requirements.txt):
fastmcp
pinecone
sentence-transformers
python-dotenv
uvicorn
httpx
fastapi
Environment Variables
Create a .env file in the project root with the following variables:
| Variable | Required | Description |
|---|---|---|
PESTS_DISEASES_RAG_URL |
✅ | Base URL of the Pests & Diseases RAG service |
GOVT_SCHEMES_RAG_URL |
✅ | Base URL of the Government Schemes RAG service |
PINECONE_API_KEY |
✅ | API key for Pinecone |
PINECONE_INDEX |
✅ | Name of the Pinecone index to query |
RAG_TIMEOUT |
❌ | HTTP timeout in seconds for RAG calls (default: 30) |
Example .env:
PESTS_DISEASES_RAG_URL= ___URL__
GOVT_SCHEMES_RAG_URL= __URL__
PINECONE_API_KEY=your-pinecone-api-key
PINECONE_INDEX=your-index-name
RAG_TIMEOUT=30
Running the Server
python alumnx_mcp_server.py
The server starts on port 9000 by default.
| Endpoint | Description |
|---|---|
/mcp |
MCP protocol endpoint (FastMCP) |
/callTool |
REST tool call endpoint |
/list-tools |
Lists all available tools |
/health |
Health check |
MCP Tools
pests_and_diseases
Query the RAG system for information about pests and diseases affecting crops.
| Parameter | Type | Required | Default | Description |
|---|---|---|---|---|
pest_name |
string | ✅ | — | Name of the pest or disease |
crop |
string | ❌ | "General" |
Crop affected by the pest or disease |
Example response:
{
"status": "success",
"information": "...",
"sources": ["..."]
}
govt_schemes
Query the RAG system for government schemes related to agriculture.
| Parameter | Type | Required | Default | Description |
|---|---|---|---|---|
scheme_type |
string | ✅ | — | Type or topic of the scheme |
state |
string | ❌ | "All India" |
State for which to retrieve schemes |
Example response:
{
"status": "success",
"information": "...",
"sources": ["..."]
}
sme_divesh
Semantic search over the SME-Divesh Pinecone namespace using all-MiniLM-L6-v2 embeddings.
| Parameter | Type | Required | Default | Description |
|---|---|---|---|---|
query |
string | ✅ | — | The search query |
top_k |
integer | ❌ | 5 |
Number of top results to return |
Example response:
{
"status": "success",
"query": "crop irrigation techniques",
"results": [
{
"score": 0.91,
"text": "...",
"source": "...",
"chunk_index": 2
}
]
}
REST API Usage
All tools are also callable via the /callTool POST endpoint:
curl -X POST http://localhost:9000/callTool \
-H "Content-Type: application/json" \
-d '{
"name": "pests_and_diseases",
"arguments": {
"pest_name": "aphids",
"crop": "wheat"
}
}'
Architecture
┌─────────────────────────────────────────┐
│ Alumnx MCP Server │
│ │
│ FastAPI App │
│ ├── /mcp ← FastMCP (MCP) │
│ ├── /callTool ← REST interface │
│ ├── /list-tools ← Tool discovery │
│ └── /health ← Health check │
│ │
│ Tools │
│ ├── pests_and_diseases → RAG HTTP call │
│ ├── govt_schemes → RAG HTTP call │
│ └── sme_divesh → Pinecone │
└─────────────────────────────────────────┘
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。
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。
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。
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。
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。
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。
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
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。