Slack

Deploy agents to Slack with streaming responses.

The Slack interface turns any Agno agent, team, or workflow into an interactive Slack app with streaming task cards, suggested prompts, and thread-based sessions.

Quick Start

slack_agent.py

from agno.agent import Agent
from agno.models.openai import OpenAIResponses
from agno.os.app import AgentOS
from agno.os.interfaces.slack import Slack

agent = Agent(name="Assistant", model=OpenAIResponses(id="gpt-5.4"))

agent_os = AgentOS(
    agents=[agent],
    interfaces=[Slack(agent=agent)],
)
app = agent_os.get_app()

uv pip install 'agno[slack]'
python -m agno.os.serve slack_agent:app --port 7777

How It Works

Concept	Behavior
Threads	Each Slack thread is a separate session with its own history
Streaming	Responses stream as live task cards with tool calls visible
@mentions	By default, responds to @mentions in channels and all DMs
Files	Users can send files; agents can upload responses

Setup

Requires a Slack App with OAuth scopes and event subscriptions. See the setup guide for step-by-step instructions.

Environment variables:

export SLACK_TOKEN="xoxb-..."
export SLACK_SIGNING_SECRET="..."

Next Steps

Full Guide

Sessions, files, SlackTools, multi-instance, and troubleshooting.

Reference

All parameters, scopes, endpoints, and events.

Examples

14 examples: teams, workflows, streaming, multimodal.

SlackTools

Give agents messaging, search, and file capabilities.