disler

disler / bowser

Public

Agentic browser automation and ui testing system — built with composable skills, subagent, command, and justfile layered architecture for repeatable, deployable browser use.

169
45
100% credibility
Found Feb 17, 2026 at 93 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Just
AI Summary

Bowser is a layered system for AI agents to automate browser interactions and perform UI testing workflows using simple stories and commands within an AI coding environment.

How It Works

1
🔍 Discover Bowser

You find this handy tool for automatically checking websites and doing browser tasks, and watch a quick video to see it in action.

2
💻 Set up your AI helper

You download a friendly AI assistant app that makes everything easy, and open your Bowser project inside it.

3
🧰 Add browser tools

You grab simple helpers that let the AI control web browsers just like you would.

4
Pick your style
🖥️
Personal tasks

Automate everyday chores like checking emails or shopping carts in your familiar browser.

📋
Group checks

Test multiple user journeys on websites, like login flows or page navigation, all in parallel.

5
🚀 Launch the magic

You give a simple instruction, and watch as smart helpers take over the browser to complete your tasks or tests automatically.

6
📸 Review the proof

Screenshots and reports pop up showing exactly what happened on each page, step by step.

✅ Mission accomplished

Your websites are validated, tasks are done, and you saved hours of manual work with reliable results.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 93 to 169 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 bowser?

Bowser is an agentic browser automation and UI testing system on GitHub, letting you drive browsers via AI agents for tasks like scraping, form filling, and validation workflows. It solves inconsistent tooling for agentic AI-powered web automation by offering both personal Chrome sessions (with your cookies and extensions) and isolated Playwright runs, all orchestrated through Just recipes and YAML-defined stories. Built for Claude Code, it supports headed or headless modes with one-command reusability.

Why is it gaining traction?

Its four-layer stack—skills for browser control, subagents for parallel scaling, commands for orchestration, and Justfile for CLI shortcuts—lets you test or deploy at any level without full setups. Developers dig the token-efficient CLI actions, step-by-step screenshots, and auto-discovery of YAML user stories for instant QA parallelism. Compared to raw Playwright or agentic browser chrome extensions, it composes personal automation (like agentic GitHub Copilot tasks) with scalable testing seamlessly.

Who should use this?

QA engineers validating UI stories across staging sites, frontend devs automating regression tests on Hacker News-style feeds or localhost apps, and AI workflow builders handling logged-in flows on Amazon or Gmail. Ideal for teams in Claude Code needing agentic browser mac/Windows setups for Perplexity-like research or Strawberry agents.

Verdict

Try it if you're deep in Claude Code and agentic browser GitHub experiments—strong docs, demo video, and Just CLI make onboarding fast despite 71 stars and 1.0% credibility score signaling early maturity. Skip for production unless you tweak for reliability; it's a solid prototype for agentic AI security and web automation prototypes.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.