mcp-client-slack
Listens for Slack @mentions and publishes them as real-time events to the KĀDI event bus, enabling event-driven architectures with low latency.
README
mcp-client-slack
MCP Server that listens for Slack @mentions and publishes them as events to the KĀDI event bus for real-time processing by template-agent-typescript.
Overview
This server is part of the KĀDI Slack bot architecture. It:
- Connects to Slack via Socket Mode to receive real-time @mention events
- Validates and publishes events to KĀDI broker using topic-based routing
- Enables event-driven architecture with <100ms latency
- Works in conjunction with mcp-server-slack (for sending replies)
Architecture
Slack @mention → Socket Mode → Event Validation → KĀDI Event Bus
↓
Topic: slack.app_mention.{BOT_USER_ID}
↓
Agent_TypeScript
↓
Claude API + Reply via mcp-server-slack
Event-Driven Architecture
This server implements a publish-subscribe pattern using KĀDI's RabbitMQ infrastructure:
- Publisher: mcp-client-slack publishes
SlackMentionEventto topicslack.app_mention.{BOT_USER_ID} - Subscriber: Agent_TypeScript subscribes to the same topic for real-time event delivery
- Benefits: Real-time processing (<100ms), no polling overhead, scalable architecture
Installation
cd C:\p4\Personal\SD\mcp-client-slack
npm install
Configuration
Create .env file:
# Slack API Credentials
SLACK_BOT_TOKEN=xoxb-your-bot-token
SLACK_APP_TOKEN=xapp-your-app-token
# Anthropic Claude API (optional, not used in this server)
ANTHROPIC_API_KEY=sk-ant-your-key
# MCP Server Configuration
MCP_LOG_LEVEL=info
# KĀDI Broker Configuration (REQUIRED for event publishing)
KADI_BROKER_URL=ws://localhost:8080
SLACK_BOT_USER_ID=U01234ABCD
Environment Variables
| Variable | Required | Description | Example |
|---|---|---|---|
SLACK_BOT_TOKEN |
✅ Yes | Slack bot user OAuth token | xoxb-123... |
SLACK_APP_TOKEN |
✅ Yes | Slack app-level token for Socket Mode | xapp-1-A... |
KADI_BROKER_URL |
✅ Yes | WebSocket URL for KĀDI broker | ws://localhost:8080 |
SLACK_BOT_USER_ID |
✅ Yes | Slack bot user ID for event topic routing | U01234ABCD |
MCP_LOG_LEVEL |
No | Logging level (debug, info, warn, error) | info |
ANTHROPIC_API_KEY |
No | Not used (kept for backward compatibility) | sk-ant-... |
Required Slack Scopes
app_mentions:read- Listen for @mentionschat:write- (handled by mcp-server-slack)
Usage
Development Mode
npm run dev
Production Mode
npm run build
npm start
As MCP Upstream (via KADI Broker)
Add to kadi-broker/mcp-upstreams.json:
{
"id": "slack-client",
"name": "Slack Event Listener",
"type": "stdio",
"prefix": "slack_client",
"networks": ["slack"],
"stdio": {
"command": "node",
"args": ["C:/p4/Personal/SD/mcp-client-slack/dist/index.js"],
"env": {
"SLACK_BOT_TOKEN": "xoxb-...",
"SLACK_APP_TOKEN": "xapp-...",
"ANTHROPIC_API_KEY": "sk-ant-..."
}
}
}
🔴 Breaking Changes
Version 2.0.0 - Event-Driven Architecture
What Changed:
- ❌ REMOVED:
get_slack_mentionsMCP tool (polling-based) - ✅ ADDED: Event publishing to KĀDI event bus (real-time)
Impact:
- Agents can no longer poll for mentions using
get_slack_mentions - Agents must subscribe to
slack.app_mention.{BOT_USER_ID}events instead
Migration Required:
- Update Agent_TypeScript to use event subscription (see Migration Guide below)
- Add
KADI_BROKER_URLandSLACK_BOT_USER_IDto environment configuration
Event Topics
Topic Pattern Standard
All KĀDI events follow the standardized topic pattern: {platform}.{event_type}.{bot_id}
{platform}: Platform identifier (slack,discord, etc.){event_type}: Event type (app_mention,mention, etc.){bot_id}: Bot unique identifier (critical for multi-bot deployments)
Why bot_id is required:
- ✅ Enables multiple bot instances in the same workspace
- ✅ Routes events only to the intended bot
- ✅ Prevents cross-bot event delivery
- ✅ Supports multi-tenant deployments
Validation: Topics are automatically validated by @agents/shared package. Invalid patterns will log warnings but still publish.
See TOPIC_PATTERN.md for complete documentation.
slack.app_mention.{BOT_USER_ID}
Published when the bot is @mentioned in Slack.
Event Payload (SlackMentionEvent):
{
id: string; // Unique mention ID (Slack event timestamp)
user: string; // Slack user ID who mentioned the bot
text: string; // Message text (with @bot mention removed)
channel: string; // Slack channel ID
thread_ts: string; // Thread timestamp for replies
ts: string; // Event timestamp from Slack
bot_id: string; // Slack bot user ID (routing identifier)
timestamp: string; // ISO 8601 datetime when event was published
}
Example:
{
"id": "1234567890.123456",
"user": "U12345",
"text": "What's the weather today?",
"channel": "C12345",
"thread_ts": "1234567890.123456",
"ts": "1234567890.123456",
"bot_id": "U01234ABCD",
"timestamp": "2025-01-16T10:30:00.000Z"
}
Integration with Agent_TypeScript
Event Subscription (Current - v2.0.0+)
Agent_TypeScript subscribes to real-time events:
import { SlackMentionEventSchema } from './types/slack-events.js';
// Subscribe to Slack mention events
client.subscribeToEvent(`slack.app_mention.${botUserId}`, async (event: unknown) => {
// Validate event payload
const validationResult = SlackMentionEventSchema.safeParse(event);
if (!validationResult.success) {
console.error('Invalid event:', validationResult.error);
return;
}
const mention = validationResult.data;
// Process mention with Claude API
// Reply using mcp-server-slack tools
});
Polling (Deprecated - v1.x)
⚠️ No longer supported - Update to event subscription:
// ❌ OLD (v1.x) - No longer works
const result = await client.getBrokerProtocol().invokeTool({
targetAgent: 'slack-client',
toolName: 'slack_client_get_slack_mentions',
toolInput: { limit: 5 },
timeout: 10000
});
Migration Guide
Upgrading from v1.x to v2.0.0
Step 1: Update Environment Variables
Add to .env:
KADI_BROKER_URL=ws://localhost:8080
SLACK_BOT_USER_ID=U01234ABCD # Get from Slack app settings
Step 2: Update Agent Code
Remove polling logic:
// ❌ Remove this
setInterval(async () => {
const result = await invokeTool({
targetAgent: 'slack-client',
toolName: 'slack_client_get_slack_mentions',
...
});
}, 10000);
Add event subscription:
// ✅ Add this
import { SlackMentionEventSchema } from './types/slack-events.js';
client.subscribeToEvent(
`slack.app_mention.${process.env.SLACK_BOT_USER_ID}`,
async (event: unknown) => {
const mention = SlackMentionEventSchema.parse(event);
// Your existing processing logic
}
);
Step 3: Update Dependencies
Install Zod for schema validation (if not already installed):
npm install zod
Step 4: Test Event Flow
- Start KĀDI broker
- Start mcp-client-slack (publisher)
- Start Agent_TypeScript (subscriber)
- Send test @mention in Slack
- Verify event received in <100ms
License
MIT
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.