CopilotKit Channels SDK
Open-source SDK that brings any AG-UI-compatible AI agent into Slack, Microsoft Teams, Discord, and Telegram with native interactive UI.
Channels SDK earns its place if your users already live in Slack or Teams: it removes provider plumbing, auth, and per-platform UI rendering, and leaves the interesting part — the agent — to you. The AG-UI framing is the real draw, since you can pick LangGraph, CrewAI, Mastra, Pydantic AI, or Google ADK and swap later rather than marrying the built-in agent. Just don't mistake it for a product: it's a plumbing layer with no analytics dashboard, and its own docs note that complex agent workflows need a significant amount of configuration. Slack Bolt alone is less machinery for a single-channel bot.
Verified 14d ago · liveness 73/100 · cite: rightaichoice.com/tools/copilotkit-channels-sdk
- Developers who already have an AG-UI-compatible agent and need it inside Slack or Teams
- Teams planning to add a second or third chat channel without rewriting agent logic
- Engineers who want approval gates and native interactive UI inside the conversation
- Shops that need managed Slack and Teams connections while keeping agent logic on their own infrastructure
- Non-developers wanting a no-code bot builder — setup is CLI, .env files, and a Node process
- Single-channel Slack bots where a direct Slack Bolt app is less machinery
- Serverless-only architectures — Channels expects a long-running process or container
We scan live Reddit threads, YouTube comments, X posts, G2 reviews and other communities — and hand you an honest verdict in under a minute.
- Honest verdict, not marketing
- Real pros & cons from real users
- Attributed quotes with receipts
3 free scans · no card needed
Skip CopilotKit Channels SDK if you want a no-code bot builder, need only a single Slack workspace (a direct Slack Bolt app is less machinery), or run serverless-only with no long-running process.
Managed Slack and Teams connections route through CopilotKit Intelligence, a separate service from the MIT-licensed SDK, so free-tier channel limits apply there rather than to the code.
The SDK itself is MIT-licensed and free to use, which suits individual engineers and small teams shipping an internal agent. The pricing question sits with CopilotKit Intelligence, the managed connection layer for Slack and Teams, whose tiers aren't published on the repo. Compare against building directly on Slack Bolt (free, one platform only) or a commercial no-code bot platform if you want a published price list before you commit.
In short
CopilotKit Channels SDK — Open-source SDK that brings any AG-UI-compatible AI agent into Slack, Microsoft Teams, Discord, and Telegram with native interactive UI. Best for Developers who already have an AG-UI-compatible agent and need it inside Slack or Teams, Teams planning to add a second or third chat channel without rewriting agent logic, Engineers who want approval gates and native interactive UI inside the conversation. Free to use.
What people actually say about CopilotKit Channels SDK — is it worth it?
We ran a structured research pass across product reviews, community discussions, and post-purchase forum threads to surface the patterns vendors won't publish themselves. Below: the recurring strengths, the hidden costs people mention most, and the cohort that consistently regrets adopting this tool.
15 mentions across 2 sources (Hacker News, Product Hunt) · researched Aug 7, 2026.
Average across the 2 sources that answered — each source counts once, not each post.
- +Unified API simplifies deploying agents across Slack and Teams.
- +Model and provider agnostic—works with your existing agent stack.
- +Clean layered architecture (adapters, ops, run loop) for sane abstractions.
- +Durable memory in Slack enables powerful stateful workflows.
- +Boilerplate reduction makes bot development noticeably faster.
- −Lack of long-term production data raises reliability questions.
- −WhatsApp, Messenger, Instagram requires paid Pro tier.
- −Documentation depth for advanced use cases is unverified.
- −Potential lock-in to CopilotKit ecosystem despite agnostic claims.
- −Support responsiveness at free tier may be limited.
- • Pro tier pricing is not publicly listed, requiring sales calls
- • Self-hosting is gated behind enterprise contracts, which may surprise some users
Viability Score
How well maintained and how widely used is CopilotKit Channels SDK? Built from what the vendor actually publishes (docs, changelog, tutorials, integrations, pricing), whether the site is live, and how much real users discuss it. How we calculate this
Last calculated: September 2026
How we score →Key Features
- Unified API to bring one agent into Slack, Microsoft Teams, Discord, and Telegram
- AG-UI-compatible agent support: LangGraph, CrewAI, Mastra, Pydantic AI, Google ADK
- CopilotKit built-in agent as an alternative to bringing your own
- Describe a message once; render as native Slack Block Kit and Teams Adaptive Cards
- Interactive UI in conversation: buttons, choices, and approval gates
- Human-in-the-loop pause for approval before the agent acts
- Streaming responses and tool calls inside the chat thread
- File handling inside channel conversations
- Managed Slack and Microsoft Teams connections through CopilotKit Intelligence
- CLI: copilotkit channels add --adapter slack and copilotkit channels status
- channels-setup skill driven by your coding agent; prints and copies a prompt
- Slack setup skill installable via npx copilotkit@latest skills install
- Live workflow fetched from copilotkit.ai/channels-guide.md at run time
- Secrets stay in .env; no CLI flag accepts a credential value
- MIT-licensed SDK on GitHub; Node.js 22+ runtime
About CopilotKit Channels SDK
CopilotKit Channels SDK is an MIT-licensed developer toolkit (roughly 912 GitHub stars) that puts an AG-UI-compatible agent inside the chat platforms your team already uses. You write the agent once, declare a Channel with `copilotkit channels add --adapter slack`, and it appears with native UI — Slack Block Kit on one side, Teams Adaptive Cards on the other — instead of you hand-rolling an adapter per platform. Your agent keeps its own tools, model, and business logic: bring your own via LangGraph, CrewAI, Mastra, Pydantic AI, or Google ADK, or use CopilotKit's built-in agent. Channels supplies the conversational surface, reading the thread, streaming a response, calling tools, handling files, rendering buttons and choices, and pausing for human approval before the agent acts. Setup is deliberately CLI-driven: `npx copilotkit@latest channels setup` installs a channels-setup skill, prints a prompt, and copies it to your clipboard for your coding agent to run, and the skill fetches the live workflow from copilotkit.ai/channels-guide.md at run time so steps stay current even if the installed skill is months old. Secrets stay in your .env; no CLI flag accepts a credential value. Managed Slack and Microsoft Teams connections route through CopilotKit Intelligence while your agent runtime stays on your own infrastructure. Node.js 22+ is required, and @copilotkit/channels ships as a tested pair with @copilotkit/runtime so you upgrade both together. This is aimed at engineers shipping agents into internal collaboration tools, not at no-code bot builders.
Behind the Verdict
The strength here is architectural honesty. Channels doesn't try to be your agent — it's a transport and rendering layer, and it says so. You keep your model, your tools, and your business logic; Channels handles the part nobody enjoys writing, which is the per-platform adapter that turns one agent message into Slack Block Kit blocks versus Teams Adaptive Cards. That's two UIs you now describe once. It also supports Discord and Telegram per the repo description, while the core material covers Slack, Teams, and Discord. The second real strength is the human-in-the-loop primitive. Putting buttons, choices, and approval gates directly in the conversation, with the agent pausing before it acts, is exactly what separates an internal agent people trust from one they mute. Streaming responses, tool calls, and file handling inside the thread make it feel native rather than bolted on. The CLI-driven setup cuts both ways. Having your coding agent drive the Slack and Intelligence consoles in your own signed-in session, with you typing the secrets and it doing the clicking, is genuinely clever — and the guide being fetched at run time from copilotkit.ai/channels-guide.md means you aren't stuck on a stale skill. But if your coding agent has no browser or computer-use tool, it will stop and ask you to add one first. That's the intended path, not a fallback, and it will surprise teams expecting a plain `npm install` experience. On weaknesses, be clear about three things. First, there's no analytics dashboard — you get the conversation as it happens, not dashboards over what the agent did. Second, the setup expects a long-running process or container; serverless-only architectures don't fit this shape. Third, documentation indicates complex agent workflows require a substantial amount of configuration, so budget real engineering time rather than an afternoon. Where it fits: developer teams with an existing AG-UI-compatible agent who need it in Slack or Teams now and may add Discord or Telegram later without rewriting agent logic. Where it doesn't: non-developers who want a no-code bot builder, single-channel Slack bots better served by a direct Slack Bolt app, and anyone who needs a published, confirmable tier list before committing — vendor pricing isn't visible on the repo.
Researching CopilotKit Channels SDK? Get your full AI stack in 60 seconds.
Free, no signup — tell us your goal and get tools matched to your budget & existing stack.
Real-world workflow fit
Concrete scenarios for the personas CopilotKit Channels SDK actually fits — and what changes day-one when you adopt it.
Runs npx copilotkit@latest channels setup, pastes the printed prompt into their coding agent, and lets it walk through the project, agent, managed Channel, and Slack provider app while they type the secrets into .env.
Outcome: The agent answers inside a Slack thread with Block Kit UI and can pause for approval before acting, with no per-platform adapter written by hand.
After shipping in Slack with copilotkit channels add --adapter slack, they declare a Teams Channel and rely on Channels to render the same described message as Adaptive Cards.
Outcome: Teams users get the same agent without any change to agent tools, model, or business logic.
Puts buttons, choices, and an approval gate directly into the conversation so the agent pauses before it acts, using the human-in-the-loop primitive rather than building an external review queue.
Outcome: The agent can stream a response, call tools, and handle files in-thread, but stops for a human decision before committing the action.
Use Cases
- Deploy a customer support assistant directly in Slack to answer common queries.
- Create a project management bot for Teams that can schedule meetings and update tasks.
- Build an internal knowledge base agent that searches documentation and replies in Slack threads.
- Integrate a data analysis assistant that can query databases and post charts to Teams.
- Add Discord as a second agent channel without rewriting your existing agent logic.
- Put an approval gate in the thread so an agent pauses before taking a risky action.
Limitations
- Setup assumes a coding agent with a browser or computer-use tool to drive the Slack and Intelligence consoles; without one it will ask you to add that first, which is the intended path rather than a fallback.
- Channels expects a long-running process or container, so serverless-only architectures don't fit the shape of the runtime.
- There is no analytics dashboard — you see the conversation as it happens, not reporting over agent activity.
- Documentation indicates the SDK may require a significant amount of configuration for complex agent workflows.
- The setup skill installed via npx is scoped to Slack; for Microsoft Teams you go through the hosted guide instead.
- Secrets must be entered by hand into .env, and @copilotkit/channels ships as a tested pair with @copilotkit/runtime, so the two need to be upgraded together.
as of 2026-09-15
Verification history
We have re-verified CopilotKit Channels SDK 4 times since . Each pass re-reads the vendor's own pages and re-checks every listed field against that evidence; passes where nothing had changed are marked as such.
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
- — re-checked, vendor evidence unchanged
Free to cite with attribution — this page re-verifies continuously.
Where the pricing makes sense
The company stage and team size where CopilotKit Channels SDK's pricing actually pencils out — and where peers do it cheaper.
The SDK itself is MIT-licensed and free to use, which suits individual engineers and small teams shipping an internal agent. The pricing question sits with CopilotKit Intelligence, the managed connection layer for Slack and Teams, whose tiers aren't published on the repo. Compare against building directly on Slack Bolt (free, one platform only) or a commercial no-code bot platform if you want a published price list before you commit.
Setup time & first value
How long it actually takes to get something useful out of CopilotKit Channels SDK — broken out by persona, not the marketing-page minute.
Developers with an AG-UI-compatible agent and a coding agent that has browser or computer-use tooling: the fastest path, since the channels-setup skill drives the Slack and Intelligence consoles in your signed-in session while you type the secrets. Teams needing Microsoft Teams instead of Slack should expect the hosted guide route rather than the Slack-scoped skill. Budget extra time if your
Switching to or from CopilotKit Channels SDK
How to bring data in from common predecessors and how to get it back out — written for the switcher, not the buyer.
- →From a hand-rolled Slack Bolt adapter: declare the Channel with copilotkit channels add --adapter slack and let Channels render Block Kit instead of maintaining the adapter.
- →From a per-platform Teams bot: describe the message once and let Channels emit Adaptive Cards rather than duplicating UI logic.
- →From CopilotKit's built-in agent to your own: swap in an AG-UI-compatible agent via the factory and keep your tools, model, and business logic.
- ↗To a direct Slack Bolt app: for a single-channel bot, drop the Channel layer and talk to Slack directly.
- ↗To another AG-UI runner: because agents are AG-UI-compatible, you can move the agent to a different framework that implements the same interface.
Integrations
Resources & Guides
Tutorials & Learning

Channels SDK:1つのエージェントをあらゆるチャネルで実行(ライブイベント)
CopilotKit

CopilotKit Channels SDK: One Agent Across Slack + Discord, What Is Actually Managed?
TechWealth Hub
YouTube returned 6 videos for “CopilotKit Channels SDK”, and we withheld 4: 4 did not mention CopilotKit Channels SDK. Showing the 2 we can prove are about CopilotKit Channels SDK.
Official links
Tools that pair well with CopilotKit Channels SDK
Common stack mates teams adopt alongside CopilotKit Channels SDK, with the specific reason each pairing earns its keep.
Plain
AI support infrastructure for B2B teams — consolidate Slack, Teams, Discord, email, and in-app channels, then orchestrate AI agents on top.
Agenta
Open-source workspace for building, evaluating, and deploying AI agents you talk to in Slack, Telegram, or a browser chat
CopilotKit
The frontend stack for agentic user experience — drop-in chat, generative UI, and shared state for React apps on any AG-UI backend
Featured Head-to-Head Comparisons
Copilotkit Channels Sdk vs Air Ai
If you're a developer building AI assistants for Slack or Teams, CopilotKit Channels SDK is your pick—freemium, code-first, and buzzing with a fresh Show HN launch. Air AI is a mission-critical defense platform (see the $450M expansion and $31M contract) for government buyers who need enterprise-grade readiness orchestration, not a messenger bot. Choose based on your domain: messaging-platform chat vs. military supply chain.
Copilotkit Channels Sdk vs Temporal Ai
If you're shipping an AI assistant into Slack or Teams today, CopilotKit Channels SDK is the faster path with its ready-made connectors and React/Vue/Svelte support. If your AI agents need to survive crashes, retries, and human-in-the-loop pauses at scale, Temporal is the battle-tested engine used by OpenAI and Replit. Choose based on where your complexity lives: channel integration or workflow reliability.
Copilotkit Channels Sdk vs Cryptohopper
Choose CopilotKit Channels SDK if you're a developer building AI agents that need to live inside Slack/Teams — you'll get a unified API and the freedom to self-host on Enterprise. Choose Cryptohopper if you're a trader wanting automated crypto bots without coding, especially with copy trading and a Strategy Designer — but be ready to pay from $24/month. They don't compete directly; your role decides.
Alternatives to CopilotKit Channels SDK
View allPlain
AI support infrastructure for B2B teams — consolidate Slack, Teams, Discord, email, and in-app channels, then orchestrate AI agents on top.
Agenta
Open-source workspace for building, evaluating, and deploying AI agents you talk to in Slack, Telegram, or a browser chat
CopilotKit
The frontend stack for agentic user experience — drop-in chat, generative UI, and shared state for React apps on any AG-UI backend
Frequently Asked Questions
Topics
Used CopilotKit Channels SDK? Help shape our editorial sentiment research.