Anthropic just shipped an OpenClaw killer called Claude Code Channels, allowing you to send messages via Telegram and Discord.

An open source autonomous AI agent OpenClaw it might just have been stolen by Anthropic.

today, Anthropic announced Claude Code Channels, connects its powerful Claude Code AI agent harness to a human user’s Discord or Telegram messaging apps, allowing them to directly message Claude Code whenever they want on the go and have it write code for them. They are official documents here.

It’s not just a new UI; This is a fundamental change in how developers interact with AI agents, moving from synchronous "ask and wait" model for asynchronous, autonomous partnership. Previously, Claude Code users were stuck interacting with a somewhat scattered (in my experience) agent harness in the Claude desktop app, the terminal or supported developer environment, and the Claude mobile app. An interaction setting called Remote Control.

Now, Anthropic offers some of the same core functionality as OpenClaw, which has led to rapid adoption among software developers and vibe coders since its November 2025 release. Austrian developer Peter Steinberger (ironically, he originally named his project "A clause" In honor of Anthropic’s own AI model, Claude, who originally ran it Anthropic sent him a cease and desist order due to potential trademark violations. Steinberger what Since it was hired by Anthropic’s competitor OpenAI.)

At the heart of OpenClaw’s appeal was its ability to have a persistent, personal AI worker that could message users 24/7 over common messaging apps like iMessage, Slack, Telegram, WhatsApp, and Discord, and send AI messages to them—not just to chat, but to email and do their work, write, and work for them. creation of all programs, apply for jobs on behalf of the user, for managing full ongoing social marketing campaigns. When the AI ​​finishes a task, it can immediately notify the human user of their preferred messaging platform.

But OpenClaw also comes with a high degree of security risk (since access to the user’s hard drive and file system or other personal data) and difficulty for non-technical users, including inspiring a wave of affiliates promising greater ease and security. NanoClaw, KiloClaw and Nvidia recently announced NemoClaw.

By giving Claude Code the same basic functionality—allowing users to send messages from the popular third-party apps Discord and Telegram, and to message them after completing a task—Anthropic effectively countered OpenClaw’s appeal and offered something it didn’t: to make the Anthropic brand and users less technical with its commitment to AI safety and security.

Technology: Model Context Protocol Bridge

At the heart of this update is the Model Context Protocol (MCP) open source A standard introduced by Anthropic in 2024. Think of MCP as a universal USB-C port for AI: it provides a standardized way for an AI model to connect to external data and tools. New "Channels" In the architecture, the MCP server acts as a two-way bridge.

When developer Claude Code starts a session --channels flag, they don’t just open a conversation; they create an election service.

Using the Bun runtime, known for its extremely fast execution of JavaScript, Claude Code controls custom plugins (currently Telegram and Discord).

When a message arrives, it is logged directly into the active session <channel> event. Claude can then use its internal tools to execute code, run tests or fix bugs, and respond to the external platform using a custom application. reply instrument.

The technical achievement here is persistence. Unlike the outdated standard web chat, a Claude Code session can now run in the background or on a persistent server (like a VPS). "ping" to take action.

How to set up Claude Code Connectors on Telegram and Discord

Building these native connectors requires Claude Code v2.1.80 or later and the Bun runtime installed on your desktop or Mac. Follow the instructions here or below.

1. Setting up Telegram

1. Create your bot: Open BotDaddy Use and in Telegram /newbot command to create a unique bot and login token.

2. Install the plugin: Run Claude Code inside your terminal: /plugin install telegram@claude-plugins-official

3. Configure the token: Run away /telegram:configure <your-token> to save your credentials.

4. Restart with channels: Exit clod and restart using the channel flag: claude --channels plugin:telegram@claude-plugins-official

5. Pair your account: DM your new bot in Telegram to get the connection code, then enter it in your terminal: /telegram:access pair <code>

2. Setting up Discord

1. Create an application: Go to the Discord Developer Portal, create "New App," and reset the bot icon to copy.

2. Activate Intentions: In the bot settings you need activate it Intent of Message Content under "Privileged Gateway Intents."

3. Install and Configure: Escape in Claude Code /plugin install discord@claude-plugins-official followed /discord:configure <your-token>.

4. Start and pair: restart with claude --channels plugin:discord@claude-plugins-official. DM your bot on Discord and use it /discord:access pair <code> end link command.

Product: From desktop "Everywhere"

The immediate practical impact is the democratization of mobile AI coding. Previously, a developer had to rely on complex self-hosted installations like OpenClaw if they wanted to check the status of a build or quickly fix it remotely from their desktop.

Installation with channels is local. A developer can create a Telegram bot through BotFather and link it to Claude Code /telegram:configure order and "couple" their account with a security code. Once configured, the phone becomes a remote control tool for the development environment.

The product also provides a "Fakechat" demo — a local-only chat interface that lets developers try it out "to push" logic on their own machines before connecting to external servers. This reflects Anthropic’s caution, "research review" approach, to ensure that developers understand the flow of events before submitting their terminals to the Internet.

Licensing: Power of Ownership of Open Standards

The licensing results of this release highlight a growing trend in the AI ​​industry: custom engines that run on open roads. Claude Code remains a proprietary product associated with Anthropic’s commercial subscriptions (Pro, Max, and Enterprise).

However, based on the open-source Model Context Protocol, Anthropic encourages the creation of a developer ecosystem. "connectors" this makes their model more useful.

While the main Claude "the brain" closed, plugins for Telegram and Discord are hosted on GitHub under the official Anthropic repositories, possibly allowing for community contributions or forks.

This strategy allows Anthropic to maintain the security and quality of the model while benefiting from the rapid innovation of the open source community – a direct challenge. "for free" but often purely the fragmented nature of open source agent frameworks.

And since it’s built on MCP, the community can build now "Connectors" For Slack or WhatsApp themselves, instead of waiting for Anthropic to send them.

Community Reactions: ‘OpenClaw Killer’

The response from users, especially AI watchers on X, was swift and decisive. The emotion was captured by the best Ejaaz (@cryptopunk7213)noted that Anthropic’s speed of delivery—text messages, thousands of MCP skills, and autonomous bug fixes in just four weeks— "crazy"

For many, this update makes native first agent frameworks obsolete. BentoBoi (@BentoBoiNFT) observation, "Claude just killed OpenClaw with this update. No need to buy a Mac Mini anymore. I say this as someone who owns a lol" It refers to the common practice of developers who purchase special hardware to run open source agents such as OpenClaw 24/7. By transferring this persistence to the Claude Code environment, Anthropic simplified the program "hardware tax" for autonomy.

AI YouTuber Matthew Berman summed up the change succinctly: "They founded OpenClaw."

The consensus among early adopters is that Anthropic has successfully embraced the most desirable features of the open source movement—multichannel support and long-term memory—while maintaining the credibility of a premier AI provider.

While Anthropic’s Claude has long been a favorite for his reasoning, he remains one "a brain in a jar"— a stateless entity that waits for the user to type before thinking. In the meantime, it has thrived by offering open source projects such as OpenClaw "always on" persistence, which allows developers to send messages to their AI from Telegram or Discord to trigger complex workflows.

Now that Anthropic has closed the gap, it’s up to users to choose which approach is best for them.

Source link