MCP Slack Daily Briefing System

An intelligent daily briefing system that automatically monitors multiple Slack channels, extracts critical messages, and generates AI-powered summaries for SRE teams.

🌟 Overview

This system uses the Model Context Protocol (MCP) to connect Claude AI with Slack, enabling automated collection and AI-powered summarization of important messages from multiple channels.

Key Features

✅ MCP-based Architecture: Uses standardized MCP protocol for AI-to-Slack communication
✅ Multi-Channel Monitoring: Aggregates messages from multiple Slack channels
✅ Intelligent Filtering: Automatically detects important messages using patterns and keywords
✅ AI-Powered Summarization: Claude generates structured daily briefings
✅ Automated Scheduling: Runs daily via systemd timers (or cron)
✅ Containerized Slack Server: Uses Podman for secure, isolated execution
✅ Threaded Slack Posts: Clean channel organization with header + threaded detail

📚 Documentation

SLACK_BRIEFING_SHOWCASE.md - Complete technical showcase and architecture
SLACK_BRIEFING_SOURCE_CODE.md - Full source code documentation (51KB+)

🚀 Quick Start

Prerequisites

Python 3.12+ with Poetry
Podman (container runtime)
Slack workspace with appropriate permissions
Anthropic Claude API access (via Claude Code)

Installation

Clone the repository

git clone https://github.com/sodoityu/mcp-slack-briefing.git
cd mcp-slack-briefing

Install Python dependencies

poetry install
# or
pip install mcp anthropic requests python-dotenv

Pull the Slack MCP container

podman pull quay.io/redhat-ai-tools/slack-mcp:latest

Configure Slack credentials
- Copy .mcp.json.example to .mcp.json
- Add your Slack session tokens (xoxc/xoxd)
- Update channel IDs in daily_briefing.py

Set up automation (optional)

# Copy systemd timer files (see cronlearn.md for details)
cp systemd/* ~/.config/systemd/user/
systemctl --user enable --now daily-briefing.timer

📖 Usage

Manual Collection

# Collect messages from the last 24 hours
poetry run python daily_briefing.py

# Custom time range (48 hours)
poetry run python daily_briefing.py 48 briefing_custom.txt

Generate Summary with Claude

Run the collection script (or wait for automated run)
Open Claude Code
Say: "Create today's daily briefing"
Review the AI-generated summary
Approve posting to your target Slack channel

Post to Slack

# Post summary to Slack
poetry run python post_summary_to_slack.py briefing_summary.txt 2026-03-16 2026-03-17 C04F0GWTD9B

🏗️ Architecture

┌─────────────────────────────────────────────────────────┐
│                  Daily Briefing System                  │
└─────────────────────────────────────────────────────────┘

Systemd Timer/Cron
       │
       ▼
run_daily_briefing.sh
       │
       ▼
daily_briefing.py ──► Slack MCP Server (Podman)
       │                      │
       │                      ▼
       │               Slack API (4 channels)
       ▼
  briefing_YYYY-MM-DD.txt
       │
       ▼
  Claude Code (Manual Review)
       │
       ▼
  Claude API (Summarization)
       │
       ▼
post_summary_to_slack.py ──► Slack Channel
       │
       ▼
  📋 Daily Briefing Posted

🛠️ Technology Stack

MCP Protocol: Standardized AI-to-service communication
Python 3.12: Core scripting language
Poetry: Dependency management
Podman: Containerized Slack MCP server
Anthropic Claude: AI-powered summarization
Systemd Timers: Modern scheduling (alternative: cron)
Bash: Automation scripts

📂 File Structure

mcp-slack-briefing/
├── README.md                          # This file
├── SLACK_BRIEFING_SHOWCASE.md         # Technical showcase
├── SLACK_BRIEFING_SOURCE_CODE.md      # Complete source code docs
├── podmanlearn.md                     # Podman workflow guide
├── cronlearn.md                       # Systemd timers guide
├── .mcp.json.example                  # MCP configuration template
├── daily_briefing.py                  # Main collection script
├── post_summary_to_slack.py           # Slack posting script
└── run_daily_briefing.sh              # Automation wrapper

⚙️ Configuration

Channel Configuration

Edit daily_briefing.py to customize monitored channels:

channels = [
    {"id": "CXXXXXXXXXX", "name": "your-channel-1"},
    {"id": "CYYYYYYYYYY", "name": "your-channel-2"},
    # Add more channels...
]

Filtering Patterns

Customize the filter_important_messages() function:

ticket_patterns = [
    r'TICKET-\d+',      # Add your ticket patterns
    r'PR #\d+',         # Pull requests
]

important_keywords = [
    'urgent', 'critical', 'incident',  # Add keywords
]

important_emojis = ['🔥', '⚠️', '🚨']  # Add emojis

Scheduling

Systemd Timer (recommended):

# Edit timer schedule
systemctl --user edit --full daily-briefing.timer

# Change this line:
OnCalendar=*-*-* 09:00:00  # Daily at 9:00 AM

Cron (alternative):

# Add to crontab
0 9 * * * /path/to/run_daily_briefing.sh >> /path/to/logs/daily_briefing.log 2>&1

🔒 Security Notes

Never commit .mcp.json with real tokens (use .mcp.json.example as template)
Store tokens in environment variables or secure vault
Use .gitignore to exclude sensitive files
Rotate Slack tokens periodically
Review and sanitize messages before posting to public channels

📊 Example Output

📋 Daily Briefing — 2026-03-16 to 2026-03-17

🔴  Executive Summary
System health stable with 3 critical incidents resolved. One ongoing upgrade
issue requires monitoring. ARO HCP experiencing intermittent DNS delays.

🔴 ITN-2026-xxxx: Customer cluster stuck in upgrading state (OHSS-xxxx)
   Status: Engineering investigating CVO logs, RCA in progress

🔵 ROSA Support
- 12 new support cases opened (8 upgrades, 3 networking, 1 authentication)
- Case escalation: SREP-xxxx (P1 cluster unavailable)

📊 Statistics: 47 messages | 3 Critical | 8 Warnings

🤝 Contributing

Contributions welcome! Please:

Fork the repository
Create a feature branch
Test thoroughly
Submit a pull request

📝 License

MIT License - feel free to use and modify for your own teams!

🙏 Acknowledgments

Red Hat AI Tools - For the Slack MCP server container
Anthropic - For Claude AI and MCP protocol
Podman Project - For secure container runtime

📧 Contact

For questions or feedback, please open an issue on GitHub.

Built with ❤️ for SRE teams managing ROSA/ARO/HCP platforms

MCP Servers