## Using the MCP server remotely
Assuming you have web based access to AI chat such as Claude - https://claude.ai, use
the following steps to connect to the remove GlyGen MCP server (https://mcp.glygen.org/mcp).
After you login to https://claude.ai
```
1. Login to https://claude.ai
2. Go to "Settings" --> "Connectors" --> "Add Custom Connector"
3. Put value "Glygen MCP Server" for the name field
4. Put value "https://mcp.glygen.org/mcp" for server
```


## Deploying the MCP server on your linux VM

### Create MCP container
Run the following command to create the MCP container
```	
$ python3 create_mcp_container.py -s dev
```

### Create service for the container
Edit /usr/lib/systemd/system/docker-glygen-mcp-dev.service and 
place the following content in it.
```
[Unit]
Description=GlyGen MCP Server Container (dev)
Requires=docker.service
After=docker.service
[Service]
Restart=always
ExecStart=/usr/bin/docker start -a running_glygen_mcp_dev
ExecStop=/usr/bin/docker stop -t 2 running_glygen_mcp_dev
[Install]
WantedBy=default.target
```

This will allow you to start/stop the container with the following commands, and ensure
that the container will start on server reboot.
```
   $ sudo systemctl daemon-reload
   $ sudo systemctl enable docker-glygen-mcp-dev.service
   $ sudo systemctl start docker-glygen-mcp-dev.service
   $ sudo systemctl stop docker-glygen-mcp-dev.service
```


### Testing the MCP server
Runn the following commands to test some of the MCP tools
```
python3 test-endpoint.py -s dev -g protein -n get_protein_summary
python3 test-endpoint.py -s dev -g site -n get_site_summary
python3 test-endpoint.py -s dev -g glycan -n get_glycan_summary
python3 test-endpoint.py -s dev -g biomarker -n get_biomarker_summary
python3 test-endpoint.py -s dev -g disease -n get_disease_summary
```