Built-in Admin Tools
The Admin Chat assistant has access to the following built-in tools for managing the Menatic AI cluster.
Cluster Information
Section titled “Cluster Information”get_dashboard_stats
Section titled “get_dashboard_stats”Returns a summary of cluster resources and workload counts.
Parameters: None
Returns:
- Total nodes, notebooks, MCP servers, models
- CPU and memory allocation summary
- GPU count
Jupyter Notebooks
Section titled “Jupyter Notebooks”list_jupyter_notebooks
Section titled “list_jupyter_notebooks”Lists all deployed Jupyter notebook instances.
Parameters: None
Returns: List of notebooks with name, status, owner, and URL.
create_jupyter_notebook
Section titled “create_jupyter_notebook”Deploys a new JupyterLab notebook instance.
Parameters:
| Parameter | Type | Required | Description |
|---|---|---|---|
name | string | Yes | Notebook name (lowercase, alphanumeric, hyphens) |
owner | string | No | Owner email or username |
Returns: Notebook name, URL, and status.
MCP Servers
Section titled “MCP Servers”list_mcp_servers
Section titled “list_mcp_servers”Lists all deployed MCP server instances.
Parameters: None
Returns: List of servers with name, status, endpoint URL.
create_mcp_server
Section titled “create_mcp_server”Deploys a new MCP server.
Parameters:
| Parameter | Type | Required | Description |
|---|---|---|---|
name | string | Yes | Server name |
image | string | Yes | Docker image |
port | integer | No | Container port (default: 8080) |
AI Models
Section titled “AI Models”list_ai_models
Section titled “list_ai_models”Lists all registered AI models (local and remote).
Parameters: None
Returns: List of models with name, type (local/remote), and status.
Example Interactions
Section titled “Example Interactions”User: How many notebooks are running?