Summarize

A single-page website summarizer. Paste a URL, get a structured brief.

Setup

pnpm install
cp .env.example .env.local

Edit .env.local:

Variable	Purpose	Default
API_URL	OpenAI-compatible base URL	http://localhost:11434/v1
API_TOKEN	API key / bearer token	—
MODEL_NAME	Chat model name	llama3.2

pnpm dev

Open http://localhost:3000.

How it works

1. The browser posts a URL to POST /api/summarize.
2. The route rate-limits by client IP: 10 requests per UTC day, stored in an in-memory map.
3. The server fetches the page, strips scripts/styles/links/images, and sends the remaining text to the configured model with the notebook system prompt.
4. The model streams a JSON object (title, summary, sections) and the page renders that fixed layout.

The in-memory limiter is per process. It resets on restart and is not shared across multiple instances.