Slack AI Streaming Agent
AI-generatedSummary
Streams AI-generated responses to a Slack thread using an AI Agent, updating the message in real-time as tokens arrive.
Inputs
- Main (required) — Workflow data items; each item provides the user prompt unless Prompt Source is set to Define Below.
- Model (required) — Connected LangChain chat model that supports streaming and tool calling (AiLanguageModel connection).
- Tools — Optional tools the agent can invoke (AiTool connection).
- Memory — Optional memory to persist conversation history across runs (AiMemory connection, max 1).
Output shape
Main
Each input item is enriched with Slack channel, thread timestamp, response text, intermediate steps, tool information, model capabilities, token count, and processing duration.
Examples
Example 1: Simple Q&A without tools or memory
Provide a Model, set Prompt Source to Define Below and write a prompt, leave Tools and Memory unconnected.
Example 2: Agent with tools and memory
Connect a Model, one or more Tools, and a Memory node; set Prompt Source to Take from Previous Node to use chatInput from prior node.