browsy-core

Zero-render browser engine for AI agents. Converts HTML into a Spatial DOM — a flat list of interactive elements with bounding boxes, roles, and states — without rendering pixels.

Usage

use browsy_core::fetch::{Session, SearchEngine};

let mut session = Session::new()?;

// Browse
session.goto("https://example.com")?;
let dom = session.dom().unwrap();

for el in dom.visible() {
    if el.role.as_deref() == Some("link") {
        println!("{}: {}", el.text.as_deref().unwrap_or(""), el.href.as_deref().unwrap_or(""));
    }
}

// Search
let results = session.search("rust web frameworks")?;
for r in &results {
    println!("{} -> {}", r.title, r.url);
}

Without networking

Disable the fetch feature for a pure HTML-to-Spatial-DOM parser:

[dependencies]
browsy-core = { version = "0.1", default-features = false }

let dom = browsy_core::parse(html, 1920.0, 1080.0);

Features

Page intelligence — 12 page types detected automatically, 13 action recipes with element IDs
CAPTCHA detection — reCAPTCHA, hCaptcha, Cloudflare Turnstile, image grids
Hidden content exposure — dropdowns, modals, accordions included with hidden: true
Session API — navigate, click, type, select, search with cookie persistence
Built-in web search — DuckDuckGo and Google
Smart deduplication — 34-42% element reduction on real sites
CSS engine — flexbox, grid, variables, calc(), @media queries
6MB binary — zero runtime dependencies

Documentation

License