Browser automation controls a web browser through code to perform tasks a person would otherwise do by hand: navigating, clicking, filling forms, and reading the resulting pages. It underpins end-to-end testing, web scraping, and agent-driven browsing.
The field spans browser-level protocols such as the Chrome DevTools Protocol and WebDriver, higher-level libraries such as Puppeteer and Playwright, and purpose-built headless engines exposed over HTTP or MCP.
🎭 Comprehensive Playwright + TypeScript automation suite for Taqelah Boutique. Includes API CRUD/mocking, visual regression, and mobile emulation with CI sharding.
🤖 新一代开源 Chrome 侧边栏自主 AI 浏览器智能体：目标驱动、长程规划、视觉空间感知与物理级操控，支持 DeepSeek 思考模型与本地离线 LLM (BYOK)
Lets Claude Code and Cursor drive Chrome. Browse your real profile: -30% tokens, -25% cost, -41% tool calls, -34% tool defs, +40% faster. Direct CDP, a11y-tree refs, server-side plan executor. MIT, no paid tier.
Автоматизация динамической многошаговой веб-формы с Playwright, MCP tool, CLI, semantic field matching, загрузкой файлов и артефактами выполнения.
A local-first AI agent work assistant built on Hermes Agent — like WorkBuddy, it does your work for you. Teach it a task once and it takes over the repetitive data entry in your old, no-API systems — and keeps working even after the UI changes. 基于 Hermes 的本地 AI Agent 工作助手。
Local-first AI browser agent and MCP server for Chrome — runs on Ollama or any cloud model. Open-source alternative to OpenAI Operator.
Browser automation for the agent era. Describe a web flow — your AI agent writes the spec and runs it. Playwright-powered, YAML-first, web-only.
Use your browser as an MCP server - connect MCP clients to read pages, run scripts, and automate tasks via a browser extension and native messaging bridge.
MCP server for WebUplink
Show your browser a task once. Get a self-healing Agent Skill any agent (Claude, OpenAI, …) can run.
This repository contains a Playwright UI test boilerplate written in TypeScript and is configured to run tests in different environments such as Development, SIT, UAT, and Production.
This repository demonstrates web automation using Jest for testing and TypeScript for development. It includes test automation scripts for web applications, focusing on testing the functionality and user interactions in the browser.
LLM-powered autonomous browser agent that plans and executes structured web workflows using deterministic Playwright automation.
Universal Interface AI – A multimodal accessibility tool built with Gemini 3 Pro. Upload a screenshot or URL to generate visual summaries, step-by-step actions, JSON automation scripts, accessibility rewrites, voice commands, and branded PDF reports.
✋ Hands for your DeepSeek Harness agent — autonomous browser operation by accessibility refs, with a permission model that follows your dsh session.
Playwright project to test page elements
Stop your web-browsing AI agent from re-exploring the same sites every run. webnav records a site's navigation once, then your agent walks routes deterministically (zero-LLM, MCP-ready) — fewer tokens, fewer steps. A reusable site map, not page compression.
Computer-use automation for legacy back-office software: an LLM discovers a workflow once, it compiles to a typed capability artifact, and replay executes it deterministically with no model in the loop — with policy, redacted evidence, and human takeover of the same live session.
Automation Project for scraping Web Content to the Google Sheet.
Web-Capability: Script-first web capabilities for AI agents. Run in-page scripts, save workflows as reusable capabilities, and generate AI-native userscripts.