Icon

Step 3 - AI Agentic System_​for one tool

<p><strong>Step 4: AI Agentic System</strong></p><p></p><p>This AI-powered workflow dynamically selects and executes the most suitable tool based on user feedback, using LLM as a reasoning engine. It includes authentication to OpenAI, user input collection, tool list scanning, and a recursive loop where the LLM determines the correct path and actives the corresponding sub-workflow.</p>
Authenticate & Connect

Authenticates and connects

to the OpenAI API.

Tool list

Scans the folder of tools, extracts their path, input/output names and description.

Step 3: AI Agentic Framework - Test Tools

User Feedback & User Data

This section provides the initial input for the retention process.

User Feedback: Collects or simulates feedback from users.

User Data: Extracts customer data

View tool call history andagent response
View Agent conversation
List all workflowtools
List Files/Folders
Table Creator
connects to LLM Agentreads datasetimports toolsoptionally imports conversation tablesends promptscalls toolsexports conversation tables and output data
Agent Prompter
Provides the OpenAIAPI credentials
Credentials Configuration
Component
OpenAI LLM Selector
Workflow to Tool
Authenticates accessto the OpenAI API
OpenAI Authenticator
Concatenate

Nodes

Extensions

Links