The Admin Chat assistant understands natural language, but this page provides example phrasings for common operations.
| Intent | Example Phrasing |
|---|
| Cluster summary | ”Give me a cluster overview” / “What’s the current status?” |
| Node count | ”How many nodes are in the cluster?” |
| Resource usage | ”What’s the CPU and memory usage?” |
| GPU status | ”Do we have any GPUs available?” |
| Intent | Example Phrasing |
|---|
| List all notebooks | ”Show me all Jupyter notebooks” / “List running notebooks” |
| Check a notebook | ”Is the ‘data-science’ notebook running?” |
| Create a notebook | ”Create a notebook named ‘ml-env’ for alice@example.com” |
| Check notebook count | ”How many notebooks are deployed?” |
| Intent | Example Phrasing |
|---|
| List servers | ”Show all MCP servers” / “What MCP servers are running?” |
| Create a server | ”Deploy an MCP server called ‘tools’ using image myorg/tools:v1” |
| Check a server | ”Is the ‘weather-mcp’ server up?” |
| Intent | Example Phrasing |
|---|
| List models | ”What AI models are available?” |
| Check model status | ”Is llama3.2 running?” |
| Find GPU models | ”Which models are using GPUs?” |
- Be specific with resource names to avoid ambiguity
- The assistant will ask for confirmation before creating resources
- If a tool call fails, the assistant will explain why and suggest alternatives
- The assistant remembers context within a conversation session — you can reference earlier messages
If a requested action fails (e.g., a notebook name is already taken, insufficient cluster resources), the assistant will explain the error in plain language and suggest corrective actions.
