Channels SDK: How to bring Your Agent to Any Channel (Slack, Microsoft Teams)
Getting any agent into messaging platforms like Slack is complex. You create a Slack app, handle its events, format messages to the platform's format, handle deliveries with retries, and so on. Taking that to production is much harder. Then, if you want to bring it into another messaging platform, you do most of the work again, and again for the next platform. To remove that manual work, we're…
The Channels SDK is an open source TypeScript library that enables any AG-UI agent to be integrated into multiple messaging platforms such as Slack, Microsoft Teams, Discord, Telegram, WhatsApp and others from a single codebase. It is built on the AG-UI protocol, allowing for decoupling of the app and the agent, and supports a wide range of AG-UI compatible agent frameworks including LangChain, Google ADK, Mastra, Pydantic AI, and Claude Agent SDK.
In addition to supporting various platforms, the SDK also offers a BuiltInAgent that can run on any model, either hosted or local (via Ollama, LM Studio, or vLLM). To get started with the Channels SDK, users will need Node.js 22+, a Slack workspace, a free CopilotKit Intelligence key, and a model API key like OpenAI. The project structure includes components for the agent, channel, and tool integration such as Notion using MCP.
The Channels UI component library provides various building blocks such as Message, Section, Header, Table, Image, Chart, Actions, and Button to help users compose rich replies that are rendered natively for each platform. To set up the Slack agent, users need to create a channel in the Intelligence dashboard, choose the platform and set its name, and generate a Slack app manifest which is then installed and configured in the Slack app.
The Bot token and Signing Secret generated in this process are required for Intelligence to post as the app and verify Slack's events. Once the channel is created, the dashboard will display the necessary information for the setup.
Written by urgent.news from Dev.to's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
