LaVague  by lavague-ai

Web agent framework for automating web processes

created 1 year ago
6,101 stars

Top 8.6% on sourcepulse

GitHubView on GitHub
Project Summary

LaVague is an open-source framework for building AI Web Agents that automate tasks based on natural language objectives. It targets developers and QA engineers seeking to streamline web interactions and testing. The framework's core benefit is its ability to translate high-level goals into executable actions using a world model and an action engine, significantly increasing web automation efficiency.

How It Works

LaVague employs a two-part architecture: a World Model that interprets objectives and current web states to generate instructions, and an Action Engine that compiles these instructions into executable code (e.g., Selenium, Playwright) and performs the actions. This modular design allows for flexibility in driver selection and customization of agent behavior.

Quick Start & Requirements

  • Install with: pip install lavague
  • Requires OpenAI API key set as OPENAI_API_KEY.
  • Supports Selenium, Playwright, and Chrome Extension drivers.
  • Quick-tour and documentation available: https://github.com/lavague-ai/LaVague

Highlighted Details

  • Automates web tasks by translating natural language objectives into executable actions.
  • Includes LaVague QA for automating test writing from Gherkin specs.
  • Offers built-in contexts, customizable configurations, a test runner, and token/cost estimation tools.
  • Supports headless agents (Selenium, Playwright), iframe handling (Selenium, Playwright), and multiple tabs (Selenium, Chrome Extension).

Maintenance & Community

  • Active development with a clear contribution process outlined via GitHub issues.
  • Community support available on Discord.
  • Roadmap available for project backlog tracking.

Licensing & Compatibility

  • License not explicitly stated in the README. Compatibility for commercial use or closed-source linking is undetermined.

Limitations & Caveats

Playwright driver's headless agent and multi-tab capabilities are marked as "coming soon." The Chrome Extension driver does not support iframe handling. Telemetry data is collected by default, which users may wish to disable for privacy.

Health Check
Last commit

6 months ago

Responsiveness

Inactive

Pull Requests (30d)
0
Issues (30d)
2
Star History
93 stars in the last 90 days

Explore Similar Projects

Feedback? Help us improve.