💬 Slack Integration

Team Collaboration · Messaging · Enterprise Gateway

Break the communication barrier between LLMs and human teams! Allow your AI agent to deliver rich-format messages, structured reports, and ops-level alerts directly to Slack workspace channels or personal DMs.

OpenClaw Team

🚀 Quick Install

Run the following command in your terminal to install:

npx clawhub install slack

📊 Stats Overview

⭐ Stars	☁️ Total Calls	👥 Active Users	🎯 Stable Version
345	1.82M	1,500	v2.0.1

🎛️ How It Works

This component is OpenClaw's purpose-built "digital communications department." By connecting to Slack's mature Bot Token workflow, the LLM can not only "read" instructions, but now has a "voice" to proactively report:

📬 Full-channel Precision Routing: Supports dual addressing modes. Broadcast data to public team groups via channel_id (e.g., #general), or lock onto a specific user_id for absolutely private point-to-point notifications.
📝 Rich Text Block Kit Adaptation: Unlike traditional plain text CLI pushes, it automatically translates LLM formatting into Slack's native Block Kit. Bold text, links, quote blocks, and bullet lists are losslessly preserved.
🧵 Thread Tracking & Context Inheritance: Supports reading and replying to specific thread_ts. During troubleshooting group conversations, AI can precisely attach diagnostic results to the original thread, keeping the channel timeline clean.
🔒 Sandboxed Permission Isolation: Uses minimal chat:write authorization scope. Ensures the bot can speak while never overstepping boundaries to tamper with or peek at the team's historical confidential information.

🧭 Typical Use Cases

🚨 Scenario 1: DevOps Incident Auto-dispatch

When your agent is running daily scheduled database backup pipelines on cloud servers and encounters Connection Timeout — the model captures the cascading error, mounts the slack component, and proactively sends highlighted error stack blocks to the #backend-ops channel, @-mentioning the on-duty ops manager.

📊 Scenario 2: Periodic Report Auto-distribution

Combined with Cron jobs and search engines like Tavily, you can instruct the agent: "Every Monday at 9:00 AM, summarize competitive intelligence and send it to the Sales Director's DMs." The LLM composes a polished analysis Markdown and pushes it through this pipeline to the executive's desktop chat — forming a complete loop.

💻 Command Reference

After installation, you can let AI call these autonomously via conversation, or manually trigger operations from the CLI:

Basic point-to-point DM notification:

clawhub execute slack channel="U01234567" message="Hello Boss, the rendering task is complete!"

Send multi-line log reports to a public channel:

clawhub execute slack channel="#server-status" \
  message="$(cat ./daily_report.log) \n---\n Sent by OpenClaw Agent"

Specify custom bot identity (requires App permissions):

clawhub execute slack channel="#marketing" \
  message="Campaign Started!" as_user=true username="Marketing AI"

🛡️ Requirements & Authentication

🔑 Slack Bot Token: Go to api.slack.com to create an app and obtain the xoxb-xxx Bot User OAuth Token.
⚙️ Environment Setup: Install the app to your workspace and pre-load the constant on the host machine: export SLACK_BOT_TOKEN="xoxb-xxxxxxxxx".
📢 Target Channel Access: To send to restricted or private channels, you must first manually Invite the Bot entity into the channel via the Slack client.

🔗 View Source on GitHub