🚀 Slack Intro Bot

Automated Daily Introduction Processing System

A sophisticated tool that extracts LinkedIn profiles from Slack introduction messages and generates formatted welcome reports. Features multi-environment support, intelligent LinkedIn extraction, and robust error handling.

🎉 NEW: Claude Code MCP Integration (November 2025)

The bot now works seamlessly with Claude Code and Zapier MCP!

Instead of running as a subprocess, Claude Code now orchestrates the entire workflow using MCP tools. Read the solution summary →

Quick Usage in Claude Code

Just ask Claude Code:

"Extract Slack introductions from the last few days"

That's it! Claude Code will handle everything. See CLAUDE_CODE_USAGE.md for details.

✨ Features

• 🎯 Smart LinkedIn Extraction: Finds LinkedIn profiles in messages or user profiles
• 🔄 Multi-Environment Support: Works in Claude Code and Cursor environments
• ⚡ Optimized Processing: Only searches profiles when LinkedIn not found in messages
• 🛡️ Robust Error Handling: Multiple timeout layers prevent hanging processes
• 📊 Comprehensive Logging: Detailed progress tracking and error reporting
• 🎨 Formatted Reports: Beautiful markdown output with embedded LinkedIn links

🚀 Quick Start

For Claude Code Environment

# Recommended: Import and run
import daily_intros
daily_intros.main()

# Or use explicit import
from daily_intros import main
main()

For Cursor Code Editor

# Auto-detect today's new introductions
python3 daily_intros.py

# Or specify date range
python3 daily_intros.py 2025-09-17 2025-09-18 2025-09-18

Expected Output

• 📄 Markdown report: daily_intros_YYYY-MM-DD.md
• 🔗 LinkedIn profiles extracted and embedded
• 📊 Processing logs with detailed progress

🔍 How It Works

Three-Phase Processing

1. 📨 Message Analysis: Searches Slack #intros channel for new introductions
2. 🔗 LinkedIn Extraction: Finds LinkedIn profiles in messages or user profiles
3. 📊 Report Generation: Creates formatted markdown reports with welcome messages

Smart LinkedIn Detection

• Primary: Extracts LinkedIn URLs from introduction message content
• Fallback: Searches user profile fields when not found in messages
• Secondary: Username-based search if user ID search fails
• Optimized: Only runs expensive profile searches when needed

Multi-Environment Support

• Claude Code: Auto-detects mcp_Zapier_* functions
• Cursor Editor: Auto-detects mcp__zapier__* functions
• Seamless: No manual configuration required

📋 Requirements

System Requirements

• Python 3.7+ (tested with 3.13)
• MCP Server with Zapier integration
• Slack Workspace with #intros channel
• API Permissions: Read access to Slack messages and user profiles

Environment Setup

• Claude Code Environment: MCP Zapier server enabled
• Cursor Code Editor: MCP Zapier server configured
• Configuration File: .env with welcome message template

Configuration

The welcome message template is stored in .env:

WELCOME_MESSAGE_TEMPLATE=Aloha {first_name}!\n\nWelcome to Lenny's podcast community!\n\nHave a wonderful day!

Output

Reports are saved to ./welcome_messages/ with secure permissions (600).

Each report includes:

• User information (name, username, LinkedIn)
• Draft welcome message
• Original introduction text
• Slack message links

Usage Options

Auto-detect (Daily Use)

python3 daily_intros.py

Automatically finds new introductions since yesterday's last processed message.

Specific Date Range

python3 daily_intros.py START_DATE [END_DATE] [OUTPUT_DATE]

Examples:

# Get all messages after 2025-09-17
python3 daily_intros.py 2025-09-17

# Get messages from Sept 18 only
python3 daily_intros.py 2025-09-17 2025-09-18

# Get Sept 18 messages, save as Sept 18 report
python3 daily_intros.py 2025-09-17 2025-09-18 2025-09-18

Parameters:

• START_DATE: Messages after this date (YYYY-MM-DD)
• END_DATE: Messages up to this date (optional)
• OUTPUT_DATE: Date for output filename (optional, defaults to today)

📁 Project Structure

slack-intro-bot/
├── 📄 README.md                    # Main documentation (this file)
├── 📄 daily_intros.py              # Entry point (backward compatible)
├── 📄 intro_extraction.py          # Dual-mode entry point
├── 📄 requirements.txt             # Python dependencies
│
├── 📁 src/                         # Source code
│   ├── daily_intros.py             # Main orchestrator
│   ├── config.py                   # Configuration management
│   ├── user_profile_search.py      # LinkedIn profile extraction
│   ├── rate_limiter.py             # API rate limiting
│   │
│   ├── 📁 dual_mode/               # Dual-mode functionality
│   │   ├── intro_extraction_api.py  # Request generation & execution
│   │   ├── claude_code_executor.py  # Claude Code executor
│   │   └── mcp_adapter.py           # Environment detection
│   │
│   └── 📁 security/                # Security components
│       ├── security_config.py       # Security configuration
│       └── security_scan.py         # Security scanning
│
├── 📁 tests/                       # Test suite
│   ├── test_config.py
│   ├── test_linkedin_extraction.py
│   ├── test_security.py
│   ├── test_dual_mode.py
│   └── run_tests.py
│
├── 📁 scripts/                     # Utility scripts
│   ├── demo_dual_mode.py           # Interactive demo
│   ├── diagnose_mcp.py             # MCP diagnostics
│   └── setup_dev.py                # Development setup
│
├── 📁 docs/                        # Documentation
│   ├── README.md                   # Documentation index
│   ├── DUAL_MODE_USAGE.md          # Dual-mode guide
│   ├── MCP_SETUP.md                # MCP setup guide
│   ├── SECURITY.md                 # Security guide
│   └── archive/                    # Historical docs
│
├── 📁 reports/                     # Generated reports
│   ├── welcome_messages/           # Daily intro reports
│   └── security/                   # Security scan reports
│
└── 📁 logs/                        # Log files

Core Components

• src/daily_intros.py: Main entry point with three-phase processing
• src/user_profile_search.py: Fallback LinkedIn extraction with timeout protection
• src/dual_mode/: Dual-mode support for Cursor IDE and Claude Code
• docs/: Comprehensive documentation

🛡️ Safety & Security

Error Handling

• Multi-layer Timeouts: 30s → 45s → 60s maximum processing time
• Guaranteed Completion: Never hangs indefinitely
• Graceful Degradation: Continues processing even if profile searches fail
• Comprehensive Logging: Detailed error reporting and progress tracking

Security Features

• Environment Variables: All secrets in .env (gitignored)
• Restricted Permissions: Output files owner-only access (600)
• No Sensitive Data: No API keys or tokens in code
• Safe Processing: No external data execution or injection

Reliability

• Signal-based Timeouts: Proper cleanup of system resources
• Exception Handling: Catches all possible error scenarios
• Fallback Mechanisms: Multiple search strategies for robustness
• Process Isolation: Individual failures don't crash the entire process

🔄 Usage Patterns

Daily Workflow

1. Morning: Open Claude Code or Cursor
2. Execute: Run the daily intros script
3. Review: Check generated markdown report
4. Complete: Process takes 30-60 seconds

Batch Processing

• Multiple Users: Handles multiple introductions per execution
• Date Ranges: Process historical introductions
• Flexible Output: Customizable report naming and formatting

Integration Ready

• API Compatible: Easy integration with other tools
• Data Export: Structured output for analytics
• Customizable: Configurable templates and processing rules

---

📚 Documentation

• 📖 README.md: Main guide (this file)
• 📋 docs/PROJECT_OVERVIEW.md: Technical documentation
• 🎯 docs/DUAL_MODE_USAGE.md: Dual-mode guide
• 🔒 docs/SECURITY.md: Security guide
• ⚡ docs/QUICK_REFERENCE.md: Quick commands
• 📚 docs/README.md: Documentation index

🤝 Contributing

This project follows software engineering best practices:

• Modular Architecture: Clean separation of concerns
• Error Resilience: Comprehensive error handling
• Documentation: Both human and LLM-optimized docs
• Testing: Robust test coverage (coming soon)
• Monitoring: Detailed logging and observability