CodingLucasLi

CodingLucasLi / Clawome

Public

Clawome - AI Agent Browser | Smart Web Browsing Tool Give your AI agent eyes and hands on any website. Open-source smart browser tool for AI agents to efficiently browse, read, and interact with web pages.

clawome.com agent ai browser-extension openclaw playwright
52
8
100% credibility
Found Feb 24, 2026 at 29 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Python
AI Summary

Clawome compresses web pages into tiny structured trees for AI agents and provides a task agent to autonomously browse and complete goals.

How It Works

1
🌐 Discover Clawome

You find a helpful tool that lets AI assistants browse websites without getting overwhelmed by messy details.

2
🚀 Launch in seconds

Run one easy command to start your dashboard and web helper right on your computer.

3
✨ See pages simplify

Open any website and watch it shrink from huge clutter to a neat, tiny outline perfect for AI eyes.

4
Pick your way
🖱️
Hands-on fun

Click around, type, and explore websites like playing a game.

🤖
Auto magic

Tell it a goal like 'find school programs' and watch it plan and browse on its own.

5
🔧 Tweak if needed

Adjust speeds, sizes, or add special tweaks for your favorite sites.

✅ AI browses perfectly

Your assistant now zips through the web, finding info fast and accurately every time.

Sign up to see the full architecture

4 more

Sign Up Free

Star Growth

See how this repo grew from 29 to 52 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 Clawome?

Clawome is an open-source Python tool that gives AI agents eyes and hands to browse, read, and interact with any web pages efficiently. It launches a Playwright browser, compresses massive live DOM trees (like 300K-token Google pages down to 3K tokens) into structured summaries preserving text, interactives, and state, then exposes a REST API for navigation, clicks, typing, and more. Pair it with a React dashboard for testing, or integrate directly into your agents.

Why is it gaining traction?

It slashes LLM token costs by 99% on benchmarks while keeping 100% visible text completeness, with pluggable compressors tuned for Google, Wikipedia, Stack Overflow, and YouTube. The v2 task agent autonomously plans subtasks, executes browser actions, self-evaluates progress, and handles errors like stuck loops—all via a simple `/api/agent/start` endpoint. Developers love the one-command `./start.sh` setup and safety rails like no form submits without explicit instruction.

Who should use this?

AI agent builders crafting web research bots, e-commerce scrapers, or data extractors that need reliable browsing without fragile CSS selectors. Ideal for backend devs prototyping LLM-driven workflows, or teams replacing Selenium with token-efficient DOM APIs. Skip if you're doing one-off scraping—use it when agents must dynamically interact across sites.

Verdict

Worth forking for agent prototypes if you're in Python/Playwright; the API and docs make it instantly usable despite 17 stars and 1.0% credibility signaling early maturity. Test the task agent on real goals first—promising but watch for edge cases in production.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.