What The Chat 💬

A modular Python package for fetching and summarizing chat history from Discord or Slack channels using AI.

example.mp4

⚠️ Beta Notice: This package is currently in beta. While core functionality works, some features may be unstable or incomplete. Use at your own discretion and expect potential breaking changes in future releases.

🌟 Overview

Transform your Discord and Slack conversations into actionable insights! This package provides:

📥 Fetch all messages from Discord/Slack channels since any date
🧵 Include messages from threads within channels
🤖 Generate intelligent summaries using local or remote LLMs
💾 Save both full chat history and summaries to files
💭 Chat interactively with your conversation history
🔧 Import into your own projects and build cool things with it (bots, web apps, etc.)

✨ Features

🔗 Versatile

• Discord and Slack channel integration
• Thread support - automatically collects messages from all threads
• User reference standardization - clean, readable output

🧠 Intelligent Summarization

• Smart summarization based on timespan:
  • 📋 Project Event Update (0-2 days)
  • 📊 Periodical Digest (3-6 days)
  • 📈 Full Project Status Summary (7+ days)
• Model flexibility - local models (Ollama) or remote (OpenAI)
• User-friendly output - replaces IDs with actual usernames

💬 Interactive Features

• Chat with your history - ask questions about conversations
• Colorful terminal output with clear visual cues
• Graceful exit handling - no more ugly tracebacks
• Export formats - markdown files with accurate timestamps

🛠️ Developer Friendly

• Modular architecture - clean separation of concerns
• Python package - easy integration into other projects
• Explicit token handling - no hidden environment dependencies
• Extensible design - perfect for building bots and web apps

🚀 Quick Start

Originally created for personal use as a CLI application, this package has been refactored to also work as an importable Python package for your own projects.

💻 Running the CLI Application

1. Clone and Install

git clone https://github.com/yourusername/what-the-chat.git
cd what-the-chat

2. Choose Installation Method

🎯 Option A: Using pixi (recommended)

pixi install

🐍 Option B: Using pip

pip install -e .

3. Setup Environment Create a .env file with your tokens:

DISCORD_TOKEN=your_discord_bot_token
SLACK_TOKEN=your_slack_bot_token
OPENAI_API_KEY=your_openai_api_key  # Only needed for remote models

📦 Installing as a Package

Add to your own projects:

pip install git+https://github.com/ulfaslak/what-the-chat.git

💡 Usage Examples

🖥️ CLI Application

🎯 Basic Usage

Get a 30-day summary with interactive chat:

# Using pixi
pixi run python apps/launch_cli.py --since-days 30 --channel 123456789012345678 --chat

# Or direct script execution
python apps/launch_cli.py --since-days 30 --channel 123456789012345678 --chat

Slack channels:

python apps/launch_cli.py --since-days 30 --platform slack --channel general --chat

✨ What happens:

1. 📥 Fetches messages from the specified channel for the last 30 days
2. 🤖 Generates an intelligent summary of the chat history
3. 💬 Starts an interactive chat session for Q&A

💾 Saving to Files

Save everything for later analysis:

python apps/launch_cli.py --since-days 30 --channel 123456789012345678 \
  --dump-file ./output --dump-collected-chat-history

📁 Creates:

• discord_history_summary_[channel]_[dates].md - AI-generated summary
• discord_history_[channel]_[dates].md - Full conversation log

🌐 Using Remote Models

Switch to powerful cloud models:

python apps/launch_cli.py --since-days 30 --channel 123456789012345678 \
  --model-source remote --model gpt-4o

🐍 Python Package Integration

Build What The Chat into your own applications:

import os
from datetime import datetime, timedelta
from what_the_chat import DiscordPlatform, SlackPlatform, SummarizationService, ChatService

# Get tokens and API keys explicitly
discord_token = os.getenv("DISCORD_TOKEN")  # or however you manage secrets
slack_token = os.getenv("SLACK_TOKEN")
openai_api_key = os.getenv("OPENAI_API_KEY")

# Discord example with explicit token
discord = DiscordPlatform(discord_token)
since_date = datetime.now() - timedelta(days=7)

# Fetch Discord messages (in an async context)
chat_history, first_date = await discord.fetch_messages_with_token(channel_id, since_date)
user_mapping = discord.get_user_mapping()

# Or Slack example with explicit token
slack = SlackPlatform(slack_token)
chat_history, first_date = slack.fetch_messages_with_token("general", since_date)
user_mapping = slack.get_user_mapping()

# Generate summary with explicit API key
summarizer = SummarizationService(model_source="remote", model="gpt-4o", api_key=openai_api_key)
summary = summarizer.generate_summary(chat_history, user_mapping)

# Start interactive chat with explicit API key
chat_service = ChatService(model_source="remote", model="gpt-4o", api_key=openai_api_key)
chat_service.start_interactive_session(chat_history, user_mapping)

# For local models, no API key needed
local_summarizer = SummarizationService(model_source="local", model="deepseek-r1-distill-qwen-7b")
local_summary = local_summarizer.generate_summary(chat_history, user_mapping)

⚙️ Command Line Arguments

Argument	Description	Required	Default
--since-days	📅 Number of days to look back	✅	-
--platform	🔗 Platform (discord/slack)	-	discord
--channel	📺 Channel ID or name	✅	-
--model-source	🧠 Model source (local/remote)	-	local
--model	🤖 Specific model name	-	deepseek-r1-distill-qwen-7b
--dump-file	💾 Save summary to file	-	-
--dump-collected-chat-history	📄 Save full chat history	-	false
--chat	💬 Start interactive chat	-	false

💬 Interactive Chat Commands

Once you're in the chat session, use these commands:

Command	Description
help	📖 Show available commands
exit, quit, q	👋 End chat gracefully
summary	📋 Generate new summary
users	👥 List all users in chat
Ctrl+C	⚡ Quick exit

🔧 Setup Guides

🎮 Discord Bot Setup

1. Create a Discord application at Discord Developer Portal
2. Create a bot for your application
3. Enable the following intents:
   • ✅ Message Content Intent
4. Generate a token for your bot
5. Invite the bot with Read Message History permissions

💼 Slack Bot Setup

1. Create a Slack app at Slack API
2. Add these OAuth scopes:
   • ✅ channels:history - read public channels
   • ✅ groups:history - read private channels
   • ✅ im:history - read direct messages
   • ✅ mpim:history - read group direct messages
   • ✅ users:read - get user information
3. Install the app to your workspace
4. Copy the Bot User OAuth Token to your .env file as SLACK_TOKEN

🧪 Testing & CI

This package includes a comprehensive test suite with automated CI/CD:

Running Tests Locally

# Run all tests
pixi run -e test pytest tests/ -v

# Run with coverage
pixi run -e test pytest tests/ --cov=what_the_chat --cov-report=term

# Run specific test file
pixi run -e test pytest tests/test_models.py -v

Automated Testing

• ✅ GitHub Actions CI - Automated testing using pixi environments
• ✅ Test Coverage - Reports sent to Codecov
• ✅ Package Installation - Verifies pixi installation and imports work
• ✅ Import Validation - Ensures all classes import correctly
• ✅ Dependency Caching - Fast CI with pixi environment caching
• ✅ Python Version Management - pixi automatically manages Python versions

Status badges in the header show real-time build and coverage status!

🤝 Contributing

Contributions are most welcome! Please feel free to submit a Pull Request.

⭐ If this project helped you, please consider giving it a star! ⭐