mirror of
https://github.com/microsoft/agent-framework.git
synced 2026-06-16 21:04:09 +08:00
Peter Ibekwe
988ef6a50e
* Initial Implementation of InvokeFunctionTool * Added unit test for InvokeFunctionTool executor. * Implemented unit and integration tests for InvokeFunctionTool. * Add sample for InvokeFunctionTool in declarative workflows. * Remove unused sample and updated comments. * Updating to official OM release with InvokeFunctionTool * Fix formatting issues. * Updated PowerFx version * Update test fixture * Cleanup - Removed unused method in InvokeFunctionToolExecutor * Update test based on PR feedback. * Update based on PR comments
56 lines
1.7 KiB
YAML
56 lines
1.7 KiB
YAML
#
|
|
# This workflow demonstrates using InvokeFunctionTool to call functions directly
|
|
# from the workflow without going through an AI agent first.
|
|
#
|
|
# InvokeFunctionTool allows workflows to:
|
|
# - Pre-fetch data before calling an AI agent
|
|
# - Execute operations directly without AI involvement
|
|
# - Store function results in workflow variables for later use
|
|
#
|
|
# Example input:
|
|
# What are the specials in the menu?
|
|
#
|
|
kind: Workflow
|
|
trigger:
|
|
|
|
kind: OnConversationStart
|
|
id: workflow_invoke_function_tool_demo
|
|
actions:
|
|
|
|
# Invoke GetSpecials function to get today's specials directly from the workflow
|
|
- kind: InvokeFunctionTool
|
|
id: invoke_get_specials
|
|
conversationId: =System.ConversationId
|
|
requireApproval: true
|
|
functionName: GetSpecials
|
|
output:
|
|
autoSend: true
|
|
result: Local.Specials
|
|
messages: Local.FunctionMessage
|
|
|
|
# Display a message showing we retrieved the specials
|
|
- kind: SendMessage
|
|
id: show_specials_intro
|
|
message: "Today's specials have been retrieved. Here they are: {Local.Specials}"
|
|
|
|
# Now use an agent to format and present the specials to the user
|
|
- kind: InvokeAzureAgent
|
|
id: invoke_menu_agent
|
|
conversationId: =System.ConversationId
|
|
agent:
|
|
name: FunctionMenuAgent
|
|
input:
|
|
messages: =UserMessage("Please describe today's specials in an appealing way.")
|
|
output:
|
|
messages: Local.AgentResponse
|
|
|
|
# Allow the user to ask follow-up questions in a loop
|
|
- kind: InvokeAzureAgent
|
|
id: invoke_followup
|
|
conversationId: =System.ConversationId
|
|
agent:
|
|
name: FunctionMenuAgent
|
|
input:
|
|
externalLoop:
|
|
when: =Upper(System.LastMessage.Text) <> "EXIT"
|