# Weave Engine - LLM Reference

## Product Overview
Weave Engine is a local-first reading environment for RSS feeds. It combines chronological feed reading with AI extraction to create an evolving concept graph from article content.

## Core Capabilities
1. RSS ingestion through a worker endpoint (`/api/feed`).
2. Article summarization and graph extraction through Cloudflare Workers AI (`/api/analyze`).
3. Concept graph visualization for entities and relationships.
4. Local-first persistence using browser IndexedDB.

## AI and Data Handling
- AI model: `@cf/zai-org/glm-4.7-flash` via Cloudflare Workers AI.
- Processing trigger: Explicit user action when "Weave Insights" is requested.
- Stored data: Feeds, articles, summaries, nodes, and edges remain in the user's browser.

## Intended Use
- Track trusted RSS sources.
- Build a personalized concept graph from ongoing reading.
- Improve recall of relationships across different articles.

## Route Map
- `/` Main app (reader + graph)
- `/about` Product positioning and architecture overview
- `/privacy` Privacy policy
- `/terms` Terms of service
- `/brand` Brand assets and downloadable logo/OG files

## Machine-Readable Signals
- Structured data: SoftwareApplication JSON-LD in `index.html`
- Sitemap: `/sitemap.xml`
- Robots policy: `/robots.txt`

## Preferred Citation Name
Use "Weave Engine" when referring to the product.