This project is a Discord bot that analyzes direct messages between two users (e.g., you and your partner), particularly focusing on:

1. Tracking and categorizing shared links
2. Searching conversation history
3. Summarizing discussions on specific topics
4. Analyzing furniture-related discussions

The project is structured to work with Val.town and has the following components:

• Database Layer: Uses Val.town's built-in SQLite database
• AI Integration: Leverages Anthropic's Claude API for analysis
• Discord Integration:
  • Bot token for fetching messages
  • Slash commands for user interactions
  • Webhook endpoint for Discord interactions
• API Endpoints: Various endpoints for manual operations and automation

To run this project on Val.town, you need to set up the following environment variables:

• DISCORD_TOKEN: Your Discord bot token
• DISCORD_CLIENT_ID: Your Discord application ID
• DISCORD_PUBLIC_KEY: Public key for verifying Discord interactions
• DISCORD_USER_ID: Your Discord user ID
• SPOUSE_USER_ID: Your partner's Discord user ID
• ANTHROPIC_API_KEY: Your Anthropic API key

1. Create a Discord application: https://discord.com/developers/applications
2. Add a bot to your application
3. Set the required permissions:
   • Read Messages/View Channels
   • Read Message History
   • Send Messages
4. Enable the Message Content Intent in the Bot section
5. Use the OAuth2 URL Generator to create an invite link with the bot scope
6. Add the bot to a server where both you and your partner are members

1. Create a new Val on Val.town
2. Copy the contents of the backend directory
3. Set up the required environment variables
4. Create a cron schedule for periodically fetching messages and processing links

• GET /health: Check if the service is running
• POST /api/interactions: Discord interactions endpoint
• POST /api/fetch-dms: Manually trigger DM fetching
• POST /api/process-links: Manually trigger link processing
• POST /api/query: Process a query about your messages
• POST /api/commands/search: Endpoint for search commands
• POST /api/register-commands: Register Discord slash commands
• GET /api/cron/fetch-dms: Cron endpoint for fetching DMs
• GET /api/cron/process-links: Cron endpoint for processing links

The bot provides several slash commands:

• /search [query]: Search your conversation history
• /links [category]: Find links shared in conversations, optionally filtered by category
• /categories: List all available link categories
• /analyze [query]: Perform a deeper analysis of your conversations

1. Clone the repository
2. Create a .env file with the required environment variables
3. Install dependencies with deno cache --reload index.ts
4. Run locally with deno run --allow-net --allow-env index.ts

A basic frontend is provided in the frontend directory. For now, we're focusing on the backend functionality, but the frontend can be extended to provide a web interface for searching and visualizing conversation data.

• The database uses Val.town's built-in SQLite database
• The code uses Deno and ESM modules compatible with Val.town
• HTTP endpoints are designed to work with Val.town's serverless model
• Discord verification is implemented according to Discord's security requirements

This project was ported from a Node.js implementation to work specifically with Val.town's environment.