Cloud AI Tool Sets User Guide
These guides cover the ready-made function-calling tool sets that attach to a
TTMSMCPCloudAI. Start with
Getting started if you have not attached one yet.
Guides
| Guide | Scope | API starting points |
|---|---|---|
| Attaching tool sets | The AI property and what registering does, how a set's tools are merged into a request, inspecting a set's declarations, withholding individual tools with Enabled, moving a set between components, teardown in either order, and deriving from a shipped set to add a tool of your own. |
TTMSMCPCloudAIToolSet, TTMSMCPCloudAITools, TTMSMCPCloudAITool |
| The four shipped tool sets | Every tool of the logger, file system, dataset, and email sets: its arguments, what it returns to the model, the component properties it reads, the ordering it expects, and the boundary it does not enforce. | TTMSMCPCloudAILogger, TTMSMCPCloudAIFileSystem, TTMSMCPCloudAIDataSet, TTMSMCPCloudAIEmail |
Related components
- Cloud AI is the component a tool set attaches to. It owns the API keys, the service selection, the prompt, and the request.
- Function calling describes the tool and parameter model a tool set is built on — read it before writing a set of your own.
- Tools publishes tools through an MCP server, so any connected client can reach them. A tool set is the opposite direction: tools your own application offers to the model it calls.