Process context and prompt to generate tool actions and results
| Property | Details |
|---|---|
| Purpose | Create intelligent conversational agents that execute workflow actions and maintain context |
| Usage | Core building block for AI assistants, chatbots, and automated workflow agents |
| Category | Intelligent Assist |
| Type | Agentic |
| Security | Immutable directives with jailbreak protection |
| Parameter | Type | Description |
|---|---|---|
| Model | modelSettings |
AI model for conversation processing and tool execution |
| Goals | textarea |
Primary objectives (immutable core directives) |
| Steps | textarea |
Mandatory process the agent must follow |
| Additional Instructions | textarea |
Extra context and behavioral guidelines |
| Data | object |
Input containing messages, thread info, and workflow tools |
| Parameter | Type | Default | Description |
|---|---|---|---|
| Thread ID | string |
- | Conversation thread identifier for continuity |
| Max Attempts | number |
10 |
Maximum tool execution attempts |
{
response: string // AI-generated conversational response
}
Returns the complete AI-generated response after processing all tools and context
The Agent Chat node provides powerful capabilities:
| Feature | Description |
|---|---|
| Immutable Goals | Primary objectives that cannot be overridden or modified |
| Mandatory Steps | Required processes the agent must strictly follow |
| Universal Standards | Built-in conversational guidelines for natural dialogue |
| Security Rules | Advanced protection against jailbreak attempts and directive overrides |
Choose the Agent Chat node when you need to:
| Aspect | Requirement | Impact |
|---|---|---|
| Goals & Steps | Must be clearly defined | Defines agent's core behavior and cannot be overridden |
| Model Selection | Choose based on complexity | Affects conversation quality and tool execution capabilities |
| Tool Definitions | Properly configured workflow steps | Required for agent to execute actions during conversations |
| Factor | Description | Optimization |
|---|---|---|
| Token Usage | Tracks prompt and completion tokens | Monitor for cost optimization |
| Max Attempts | Prevents infinite loops | Set appropriate limits based on workflow complexity |
| Message Processing | Handles text, images, data URLs | Ensure proper input validation |
| Security Feature | Protection | Benefit |
|---|---|---|
| Jailbreak Prevention | Advanced security rules | Maintains agent integrity |
| Directive Protection | Immutable core objectives | Ensures consistent behavior |
| Error Handling | Graceful failure recovery | Reliable operation under all conditions |
Goals: Provide helpful, accurate customer assistance with professional standards
Steps: Gather context Troubleshoot Escalate when needed Follow up
Tools: Knowledge base search, ticket creation, user lookup
Goals: Help users understand and analyze their data through natural conversation
Steps: Clarify requirements Perform analysis Visualize results Provide insights
Tools: Database queries, chart generation, statistical analysis
Last updated: 2025-11-28 - revision 697b76812