The goal of this Val is to receive webhooks from different notification sources (like Sentry or Linear) and pass that information through AI to determine which channels are most appropriate to notify. There's a bit of an AI spin to this and a bit of hard-coded prompting required to help guide the notifications to their right channels. We definitely want those notifications to be delivered quickly. The notifications that actually get delivered should contain very precise messages that say:

• this is what was opened
• this is what the links are according to Slack's format
• this is a summary for the intended audience which is based on the channel name

You'll have a list of channels that are hard-coded with their descriptions for what is discussed inside them. Those channels may be their email addresses or something. Maybe we start there with just email addresses and sending via email.

• CONFIGURATION.ts - Channel definitions, webhook config, and AI prompts
• linear-webhook.ts - HTTP webhook handler for Linear notifications
• main.tsx - (Reserved for future use)

1. Set environment variables:

   • LINEAR_WEBHOOK_SECRET - Your Linear webhook signing secret
   • OPENAI_API_KEY - OpenAI API key (handled by Val Town's OpenAI integration)

2. Configure your Linear webhook to point to the linear-webhook.ts endpoint

3. Update the channel email addresses in CONFIGURATION.ts to match your organization

• ✅ Linear webhook integration with signature verification
• ✅ AI-powered notification triage using GPT-4
• ✅ Email delivery to appropriate channels
• ✅ Configurable channel definitions with descriptions and keywords
• 🚧 Sentry integration (planned)
• 🚧 Slack integration (planned)