onorbumbum

onorbumbum / aslan-browser

Public

A native macOS browser built for AI agents. WKWebView + Unix socket + JSON-RPC.

onorbumbum.github.ioaslan-browser
13
0
100% credibility
Found Feb 22, 2026 at 10 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Python
AI Summary

Aslan Browser is a native macOS web browser designed for AI agents, providing fast access to page content via an accessibility tree and screenshots through a simple Python interface.

How It Works

1
📰 Discover Aslan Browser

You hear about a smart, lightweight browser made just for AI helpers on your Mac, perfect for letting them surf the web without heavy downloads.

2
📥 Download the app

Grab the ready-made app bundle for your Mac from the releases page – it works on both new and older Apple computers.

3
⚙️ Launch the helper

Start the app in the background so it's ready and waiting, invisible but fully powered up.

4
🤖 Link your AI assistant

Share simple instructions with your AI so it knows how to visit pages, spot buttons and forms, and take actions like a human user.

5
🌐 Watch AI explore

Your AI goes to websites, gets a quick list of clickable items and text, fills forms, clicks links, and snaps pictures of pages.

6
📱 Get smart page insights

Instead of overwhelming details, your AI receives a tidy summary of what's interactive on the page, saving time and effort.

✅ AI handles web tasks

Now your AI automates browsing chores like research, logins, or shopping effortlessly, fast and reliably on your Mac.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 10 to 13 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is aslan-browser?

Aslan browser is a native macOS app wrapping WKWebView for AI agent control via Unix socket and JSON-RPC. The zero-dep Python SDK handles navigation, accessibility tree extraction, element clicks/fills by @eN refs, screenshots, multi-tab sessions, and cookies. Ditch full DOM dumps for compact trees that slash LLM tokens 10-100x on complex pages.

Why is it gaining traction?

No 500MB Chrome downloads—runs native macOS code with sub-2ms JS evals and sub-30ms screenshots via local IPC. Python CLI and async/sync APIs make scripting github native client tasks dead simple, outperforming Puppeteer on memory and speed for macOS native app automation. Benchmarks prove it; agents loop faster without HTTP overhead.

Who should use this?

AI builders automating web tasks on macOS, like feeding a11y trees to LLMs for form-filling or research agents. Suits remote desktop macOS native setups, github native script prototypes, or devs scripting native macOS containers without Selenium bloat.

Verdict

Promising github native alpha (10 stars, 1.0% credibility) with stellar docs and pre-built binaries, but light tests signal early days—enable github native automerge not enabled in repo settings for polish. Use for macOS 14+ agent prototypes now.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.