The Conversation node enables intelligent dialogue management and AI-powered conversation handling within automation workflows.

Overview

The Conversation node enables intelligent dialogue management and AI-powered conversation handling within automation workflows. It supports a range of actions for interacting with large language models, collecting structured user inputs, crawling web content, executing prompts, integrating with MCP tools, and publishing structured responses — making it a core building block for agentic and conversational automation patterns.

Call LLM Model

Sends a conversation history to a large language model and retrieves an AI-generated response. Supports streaming, tool-use declarations, and session continuity.

Key Input Fields:

• Model ID — The identifier for the LLM to invoke

• System Prompt — Instructions that set the model's behavior and persona

• Session ID — Maintains conversation continuity across multiple turns

• Stream (Boolean) — Enables streaming responses for real-time output

• Tools — List of tool definitions available to the model

• Messages — Array of prior conversation messages (role + content)

• Parameters — Model parameters such as temperature, max tokens, top-p

Output: Model response text, tool call requests (if any), usage statistics, session state.

Call Interface

Invokes a named automation interface, allowing dynamic routing of the conversation to different downstream automations based on conditions.

Key Input Fields:

• Case ID — Identifier for the current conversation or support case

• Automation ID — The target automation to invoke via this interface

Output: Result from the invoked automation, passed back into the current workflow.

Collect Slots

Gathers structured data from users through a slot-filling dialogue pattern, with configurable retry behavior and auto-detection options.

Key Input Fields:

• Skip Auto Detection (Boolean) — When true, disables automatic slot filling from context

• Retries — Number of times to re-prompt the user for missing slot values

• Define Slots — Schema of slots to collect: name, type, prompt, validation rules

Output: Filled slot values as structured data, ready for downstream steps.

Crawl Website

Fetches and extracts textual content from a website URL for use as context in conversational or AI-processing workflows.

Key Input Fields:

• Website URL — The URL of the page or site to crawl

Output: Extracted page text, page title, metadata, and crawl status.

Execute Prompt

Runs a pre-configured prompt template against a specified model, with support for dynamic inputs and versioned prompts.

Key Input Fields:

• Prompt — The prompt template to execute

• Inputs — Dynamic variables to inject into the prompt template

• Model — The LLM model to use for execution

• Prompt Version — Specific version of the prompt to run, enabling A/B testing or rollback

Output: Model-generated text, token usage, model metadata.

MCP

Integrates with Model Context Protocol (MCP) tool servers, enabling the automation to use external tools declared in an MCP-compatible format.

Key Input Fields:

• Content List — List of content items (text, images, tool results) to pass as context

• Annotations — Metadata annotations attached to content items

• Is Error (Boolean) — Indicates whether the MCP content represents a tool error

Output: MCP-formatted tool response, ready to pass back to the Call LLM Model action.

Publish Response

Finalizes and sends a structured response to the end user, with support for ending the conversation, attaching files, and including call-to-action elements.

Key Input Fields:

• End Conversation (Boolean) — When true, closes the conversation session after sending the response

• Response Structure — The formatted response body (text, markdown, or structured JSON)

• Attachments — Files or media to include with the response

• CTA — Call-to-action buttons or links to embed in the response

Output: Delivery confirmation, response metadata, session closure status.

Tip: Use Collect Slots before Call LLM Model to ensure all required parameters are gathered before invoking the model, reducing unnecessary LLM calls for missing information.

Notes

Keep the following in mind when using Conversation by UnifyApps:

• Use Execute Prompt for single-step AI tasks; reserve the full agent flow (Collect Slots → Publish Response) for multi-turn conversations that need structured input gathering.

• The MCP action connects to Model Context Protocol servers; confirm the server is reachable and required tools are registered before testing.

• Crawl Website respects robots.txt by default; verify the target site permits automated access before using this action in production.

• When using Call LLM Model, pass only the context needed for the model to respond accurately; large context windows increase both latency and cost.

• Pair Publish Response with a Callable trigger when the automation needs to return a formatted AI reply to the caller.