get_instance_types

Retrieve available instance types for cloud providers to optimize resource selection for ML model deployment. Supports AWS, GCP, Azure, and more.

Instructions

Get available instances types for a cloud provider

Input Schema

TableJSON Schema

Name	Required	Description	Default
`cloud_provider`	Yes	The cloud provider to get instance types for

Implementation Reference

src/brev_mcp/tools.py:10-21 (handler)

MCP tool handler function that validates the cloud_provider argument, fetches instance types using get_provider_instance_types, and formats the response as TextContent.

async def get_instance_types_tool(args: dict[str, str]) -> TextContent:
    if "cloud_provider" not in args:
        raise ValueError("cloud_provider argument is required for get_instance_types tool")

    cloud_provider = CloudProvider(args["cloud_provider"])
    instance_types = await get_provider_instance_types(cloud_provider)
    return [
        TextContent(
            type="text", 
            text=instance_types
        )
    ]

src/brev_mcp/tools.py:42-51 (schema)

Input schema definition for the get_instance_types tool, specifying the required cloud_provider parameter with enum values from CloudProvider.

inputSchema={
    "type": "object",
    "properties": {
        "cloud_provider": {
            "description": "The cloud provider to get instance types for",
            "enum": [provider.value for provider in CloudProvider]
        }
    },
    "required": ["cloud_provider"]
}

src/brev_mcp/tools.py:38-54 (registration)

Registration of the get_instance_types tool in the tool_models dictionary, linking the tool schema, description, and handler function.

"get_instance_types": ToolModel(
    tool=Tool(
        name="get_instance_types",
        description="Get available instances types for a cloud provider",
        inputSchema={
            "type": "object",
            "properties": {
                "cloud_provider": {
                    "description": "The cloud provider to get instance types for",
                    "enum": [provider.value for provider in CloudProvider]
                }
            },
            "required": ["cloud_provider"]
        }
    ),
    call_tool=get_instance_types_tool
),

src/brev_mcp/instance_types.py:16-32 (helper)

Helper function that retrieves all instance types from the API, filters and selects those for the given provider, and returns a formatted JSON string.

async def get_provider_instance_types(provider: CloudProvider)-> str:
    try:
        all_instance_types_obj = await get_instance_types()
        instance_types = filter_instance_types(all_instance_types_obj)
        for cloud_provider, instance_type_list in instance_types.items():
            logger.info(f"Number of instance types for {cloud_provider.value}: {len(instance_type_list)}")
        if provider not in instance_types:
            raise ValueError(f"Provider {provider.value} not found in instance types")
        instance_type_dicts = [
            instance_type.model_dump(exclude_none=True) 
            for instance_type in instance_types[provider]
        ] 
        return json.dumps(instance_type_dicts, indent=2)
    except Exception as e:
        logger.error(f"Error getting instance types: {str(e)}")
        raise RuntimeError(f"Failed to get instance types: {str(e)}")

src/brev_mcp/api.py:13-32 (helper)

Core helper function that performs the HTTP API call to Brev's endpoint to fetch all available instance types, validates the response, and returns the parsed object.

async def get_instance_types() -> AllInstanceTypeObj:
    access_token = get_acess_token() 
    org_id = get_active_org_id()
    try:
        async with httpx.AsyncClient(timeout=httpx.Timeout(25.0)) as client:
            response = await client.get(
                f"{BASE_API_URL}/instances/alltypesavailable/{org_id}",
                headers={
                    "Authorization": f"Bearer {access_token}",
                    "Content-Type": "application/json"
                },
            )
            response.raise_for_status()
            data = response.json()
            all_instance_types_obj = AllInstanceTypeObj.model_validate(data)
            return all_instance_types_obj
    except ValidationError as e:    
        raise RuntimeError(f"Failed to validate instance types: {str(e)}")
    except Exception as e:
        raise RuntimeError(f"Failed to get instance types: {str(e)}")

Brev