Chan283/classifier_mcp_server
3.2
If you are the rightful owner of classifier_mcp_server and would like to certify it and/or have it hosted online, please leave a comment on the right or send an email to dayong@mcphub.com.
The Classifier MCP Server is a sophisticated system designed to facilitate seamless communication between clients and databases using the Model Context Protocol (MCP).
Classifier MCP Server
Steps to run:
1. Spin up neo4j and postgres docker containers:
docker run -d --name pgdb -e POSTGRES_PASSWORD=secret -e POSTGRES_DB=mcp_db -p 5432:5432 postgres:15
docker run -d --name neo4jdb -e NEO4J_AUTH=neo4j/test12345 -p 7474:7474 -p 7687:7687 neo4j:5
2. Install Requirements:
pip install -r requirements.txt
3. Run the mcp server:
cd mcp_server
uvicorn main:app --reload --port 8000
4. Run the mcp client:
cd mcp_client
python client.py
5. Execution
Type prompt in client. Based on this openai will decide whether to access neo4j db or postgres db