browser-use/browser-use
Browser-use is a project that enables AI agents to control and automate web browsers, allowing users to perform complex tasks through natural language commands and AI-driven browser interactions.
Awesome AI Agents › Browser Automation Tools
reworkd/bananalyzer
Banana-lyzer is an open-source AI agent evaluation framework specifically designed for web tasks, utilizing Playwright for browser automation. The project addresses the challenges of evaluating AI agents on web tasks, such as the dynamic nature of websites, latency issues, and anti-bot protections. It provides a reliable system to save and deploy historic or static snapshots of websites to ensure consistent evaluation over time. The framework supports a diverse dataset of websites across various industries and use cases to help AI agents generalize better. It focuses on structured and direct information retrieval from websites, unifying valuable web task datasets and evaluations like Mind2Web and WebArena into a single repository. The tool operates as a command-line interface (CLI) that runs a set of evaluations against example websites defined in a JSON schema similar to Mind2Web and WebArena. These examples include metadata such as agent goals, expected outputs, and snapshots of URLs saved as MHTML files to prevent changes over time. The CLI dynamically constructs a pytest test suite to run these examples against user-defined agents. Users implement the AgentRunner interface to define their agents, which interact with the provided browser context and examples. Banana-lyzer supports different test intents, including listing pages where agents scrape detail page links, detail pages where agents retrieve specific JSON information, and listing_detail pages where all required information is on the current page. Additional tags like pagination indicate tests that require fetching data across multiple pages. The project is actively developed, with plans to support more complex evaluation methods, multi-step navigation, and integration of additional datasets. The repository also includes a FastAPI server to expose example data via an API, facilitating easier access and integration. The project encourages contributions, providing tools to add new examples by saving web pages as MHTML files using Playwright. The roadmap outlines future enhancements such as multi-page examples, in-page actions, lag and bot detection emulation, and improved test visualization. Overall, Banana-lyzer is a comprehensive framework for evaluating AI agents on web-based tasks, emphasizing reproducibility, diversity, and structured data extraction.
https://github.com/reworkd/bananalyzer
Also in Browser Automation Tools
CloakHQ/CloakBrowser
CloakBrowser provides a stealthy Chromium binary with C++ source-level fingerprint patches to pass bot detection, serving as a drop-in Playwright/Puppeteer replacement.
jackwener/OpenCLI
OpenCLI converts any website or local tool into a command-line interface, enabling both human interaction and AI agent automation through logged-in browser sessions.
browserbase/stagehand
Stagehand is a production-ready AI web browsing framework that combines code and natural language to enable reliable and extensible browser automation.
browser-use/web-ui
A web-based user interface, powered by Gradio, for interacting with browser automation agents, supporting various LLMs and custom browser configurations.
citrolabs/ego-lite
ego lite is a specialized browser designed for AI agents to perform web automation in parallel with human browsing, offering isolated workspaces and enhanced page snapshots.
jo-inc/camofox-browser
Camofox-browser is a stealth headless browser server, powered by Camoufox (a Firefox fork), designed for AI agents to bypass bot detection and Cloudflare, offering a Puppeteer/Playwright alternative.
microsoft/fara
Fara-7B is a compact and efficient agentic small language model by Microsoft designed to perform multi-step computer tasks through visual interaction, achieving state-of-the-art performance on web automation benchmarks while enabling on-device deployment for privacy and low latency.