
Example Agent for TotalAgility
Example Agent for TotalAgility Overview
Features
This project provides an example TotalAgility package containing a template for a managing or planning agent, which in turn can call other worker or micro agents, as required, to complete its goal.
The agent desgin makes use of standardised input and output interfaces for Agent or Custom LLM type processes, meaning the managing agent can call any worker agent without needing to know specific details of how each worker agent is implemented.
The pattern uses a registry of available worker agents, which are used in the planning step of the managing agent to describe the capabilities of each worker agent.
This overall pattern is similar to "tool use" via MCP, and indeed the managing agent or worker agents can be called via an MCP Proxy using TotalAgility's OpenAPI REST interface.
The package was created using TotalAgility 8.1.0.
Benefits
This package can be imported to provide a template for a managing or planning agent. It can be configured for use in automation (i.e. as a top level automation agent) or as a chat engine for a "Generative AI Chat Control" in a TotalAgility Form.
It can also be used with the TotalAgility-MSTeams-Agent to add a Teams front end.
Technical Details
Inputs
Standard TotalAgility agent inputs: prompts and/or documents.
Outputs
Returns text based on the goal/task given to the agent, and the steps the agent followed to achieve that goal.

