A lightweight Node.js service that extracts and summarizes text content from URLs using modern LLM-based or heuristic summarization pipelines. Designed to be simple, fast, and easy to embed into other services or automation workflows.
- 🔗 Accepts raw URLs as input
- 🧠 Generates concise summaries of web content
- ⚡ Fast HTTP API (Node.js / TypeScript-ready)
- 🧩 Simple integration with bots, dashboards, and pipelines
- 🪶 Minimal dependencies and straightforward architecture
- 🔄 Easily extendable to swap summarization engines (LLM / extractive / hybrid)
git clone https://github.com/gusvendegna/sumArIzer.gitcd sumArIzernpm install
Start the server:
npm run dev
or
npm start
Server will typically run on:
http://localhost:3000
API: Summarize a URL Request POST /summarize Content-Type: application/json { "url": "https://example.com/article" } Response { "url": "https://example.com/article", "summary": "A concise summary of the article content.", "success": true }
Fetches HTML content from the provided URL
Extracts readable text (removing boilerplate / markup)
Passes cleaned text into a summarization pipeline
Returns a condensed summary response via HTTP API
Client → Express API → HTML Fetcher → Text Extractor → Summarizer → Response
You can configure behavior via environment variables:
- Node.js (Hono)
- TypeScript
- Ollama
There was very minimal use of AI during development of this project. I asked ChatGPT about concepts that I don't remember, but every single line of code was written from the dome or copied & pasted from a doc page.
AI was, however used for:
- This README.md
- Refining the prompts for summarization and severity classification (slop in, slop out!)
That's it.
MIT