[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"project-72048":3},{"id":4,"name":5,"fullName":6,"owner":7,"repo":5,"description":8,"homepage":9,"htmlUrl":10,"language":11,"languages":10,"totalLinesOfCode":10,"stars":12,"forks":13,"watchers":14,"openIssues":15,"contributorsCount":16,"subscribersCount":16,"size":16,"stars1d":17,"stars7d":17,"stars30d":18,"stars90d":16,"forks30d":16,"starsTrendScore":19,"compositeScore":20,"rankGlobal":10,"rankLanguage":10,"license":21,"archived":22,"fork":22,"defaultBranch":23,"hasWiki":24,"hasPages":22,"topics":25,"createdAt":10,"pushedAt":10,"updatedAt":32,"readmeContent":33,"aiSummary":34,"trendingCount":16,"starSnapshotCount":16,"syncStatus":35,"lastSyncTime":36,"discoverSource":37},72048,"LaVague","lavague-ai\u002FLaVague","lavague-ai","Large Action Model framework to develop AI Web Agents","https:\u002F\u002Fdocs.lavague.ai\u002Fen\u002Flatest\u002F",null,"Python",6366,573,53,93,0,4,26,12,76.88,"Apache License 2.0",false,"main",true,[26,27,28,29,30,31],"ai","browser","large-action-model","llm","oss","rag","2026-06-12 04:01:03","\u003Cp align=\"center\">\r\n  \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Flavague-ai\u002FLaVague\u002Fstargazers\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Flavague-ai\u002FLaVague.svg?style=for-the-badge\" alt=\"Stargazers\">\u003C\u002Fa>\r\n  \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Flavague-ai\u002FLaVague\u002Fissues\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fissues\u002Flavague-ai\u002FLaVague.svg?style=for-the-badge\" alt=\"Issues\">\u003C\u002Fa>\r\n  \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Flavague-ai\u002FLaVague\u002Fnetwork\u002Fmembers\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fforks\u002Flavague-ai\u002FLaVague.svg?style=for-the-badge\" alt=\"Forks\">\u003C\u002Fa>\r\n  \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Flavague-ai\u002FLaVague\u002Fgraphs\u002Fcontributors\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fcontributors\u002Flavague-ai\u002FLaVague.svg?style=for-the-badge\" alt=\"Contributors\">\u003C\u002Fa>\r\n\u003C\u002Fp>\r\n\u003C\u002Fbr>\r\n\r\n\u003Cdiv align=\"center\">\r\n  \u003Cimg src=\"docs\u002Fassets\u002Flogo.png\" width=140px: alt=\"LaVague Logo\">\r\n  \u003Ch1>Welcome to LaVague\u003C\u002Fh1>\r\n\r\n\u003Ch4 align=\"center\">\r\n \u003Ca href=\"https:\u002F\u002Fdiscord.gg\u002FSDxn9KpqX9\" target=\"_blank\">\r\n    \u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FDiscord-5865F2?style=for-the-badge&logo=discord&logoColor=white\" height='35px' alt=\"Join our Discord server!\">\r\n  \u003C\u002Fa>\r\n  \u003Ca href=\"https:\u002F\u002Fdocs.lavague.ai\u002Fen\u002Flatest\u002F\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002F📄-docs-000000?style=for-the-badge&colorA=09c&colorB=555\" height='35px' alt=\"Docs\">\u003C\u002Fa>\r\n\u003C\u002Fh4>\r\n  \u003Cp>A Large Action Model framework for developing AI Web Agents\r\n\u003C\u002Fp>\r\n\u003Ch1>\u003C\u002Fh1>\r\n\u003C\u002Fdiv>\r\n\r\n## LaVague: Web Agent framework for builders\r\n\r\nLaVague is an open-source framework designed for developers who want to create AI Web Agents to automate processes for their end users.\r\n\r\nOur Web Agents can take an objective, such as \"Print installation steps for Hugging Face's Diffusers library,\" and generate and perform the actions required to achieve the objective.\r\n\r\nLaVague Agents are made up of:\r\n\r\n- A World Model that takes an objective and the current state (aka the current web page) and outputs an appropriate set of instructions.\r\n- An Action Engine which “compiles” these instructions into action code, e.g., Selenium or Playwright & executes them\r\n\r\n\r\n### LaVague QA: Dedicated tooling for QA Engineers\r\n**🌊 Built on LaVague**\r\n\r\nLaVague QA is a tool tailored for QA engineers leveraging our framework. \r\n\r\nIt allows you to automate test writing by turning Gherkin specs into easy-to-integrate tests. LaVague QA is a project leveraging the LaVague framework behind the scenes to make web testing 10x more efficient.\r\n\r\nFor detailed information and setup instructions, visit the [LaVague QA documentation](https:\u002F\u002Fdocs.lavague.ai\u002Fen\u002Flatest\u002Fdocs\u002Flavague-qa\u002Fquick-tour\u002F).\r\n\r\n## 🚀 Getting Started\r\n\r\n### Demo\r\n\r\nHere is an example of how LaVague can take multiple steps to achieve the objective of \"Go on the quicktour of PEFT\":\r\n\r\n\u003Cp align=\"center\">\r\n  \u003Cimg src=\".\u002Fdocs\u002Fassets\u002Fdemo_agent_hf.gif\" alt=\"Demo for agent\">\r\n\u003C\u002Fp>\r\n\r\n### Hands-on \r\n\r\nYou can do this with the following steps:\r\n\r\n1. Download LaVague with:\r\n\r\n```bash\r\npip install lavague\r\n```\r\n2. Use our framework to build a Web Agent and implement the objective:\r\n\r\n```python\r\nfrom lavague.core import  WorldModel, ActionEngine\r\nfrom lavague.core.agents import WebAgent\r\nfrom lavague.drivers.selenium import SeleniumDriver\r\n\r\nselenium_driver = SeleniumDriver(headless=False)\r\nworld_model = WorldModel()\r\naction_engine = ActionEngine(selenium_driver)\r\nagent = WebAgent(world_model, action_engine)\r\nagent.get(\"https:\u002F\u002Fhuggingface.co\u002Fdocs\")\r\nagent.run(\"Go on the quicktour of PEFT\")\r\n\r\n# Launch Gradio Agent Demo\r\nagent.demo(\"Go on the quicktour of PEFT\")\r\n```\r\n\r\nFor more information on this example and how to use LaVague, see our [quick-tour](https:\u002F\u002Fdocs.lavague.ai\u002Fen\u002Flatest\u002Fdocs\u002Fget-started\u002Fquick-tour\u002F).\r\n\r\n> Note, these examples use our default OpenAI API configuration and you will need to set the OPENAI_API_KEY variable in your local environment with a valid API key for these to work.\r\n\r\nFor an end-to-end example of LaVague in a Google Colab, see our [quick-tour notebook](https:\u002F\u002Fcolab.research.google.com\u002Fgithub\u002Flavague-ai\u002Flavague\u002Fblob\u002Fmain\u002Fdocs\u002Fdocs\u002Fget-started\u002Fquick-tour-notebook\u002Fquick-tour.ipynb)\r\n\r\n## Key Features\r\n\r\n- ✅ [Built-in Contexts](https:\u002F\u002Fdocs.lavague.ai\u002Fen\u002Flatest\u002Fdocs\u002Fget-started\u002Fcustomization\u002F) (aka. configurations)\r\n- ✅ [Customizable configuration](https:\u002F\u002Fdocs.lavague.ai\u002Fen\u002Flatest\u002Fdocs\u002Fget-started\u002Fcustomization\u002F)\r\n- ✅ [A test runner](https:\u002F\u002Fdocs.lavague.ai\u002Fen\u002Flatest\u002Fdocs\u002Fget-started\u002Ftesting\u002F) for testing and benchmarking the performance of LaVague\r\n- ✅ A [Token Counter](https:\u002F\u002Fdocs.lavague.ai\u002Fen\u002Flatest\u002Fdocs\u002Fget-started\u002Ftoken-usage\u002F) for estimating token usage and costs\r\n- ✅ [Logging tools](https:\u002F\u002Fdocs.lavague.ai\u002Fen\u002Flatest\u002Fdocs\u002Fget-started\u002Fcustomization\u002F)\r\n- ✅ An optional, interactive [Gradio interface](https:\u002F\u002Fdocs.lavague.ai\u002Fen\u002Flatest\u002Fdocs\u002Fget-started\u002Fgradio\u002F)\r\n- ✅ [Debugging tools](https:\u002F\u002Fdocs.lavague.ai\u002Fen\u002Flatest\u002Fdocs\u002Fget-started\u002Fcustomization\u002F)\r\n- ✅ [A Chrome Extension](https:\u002F\u002Fdocs.lavague.ai\u002Fen\u002Flatest\u002Fdocs\u002Fget-started\u002Fdocs-chrome\u002F)\r\n\r\n## Supported Drivers\r\n\r\nWe support three Driver options:\r\n\r\n- A Selenium Webdriver\r\n- A Playwright webdriver\r\n- A Chrome extension driver\r\n\r\nNote that not all drivers support all agent features:\r\n\r\n| Feature                  | Selenium  | Playwright       | Chrome Extension                     |\r\n|--------------------------|-----------|------------------|--------------------------------------|\r\n| Headless agents    | ✅ | ⏳ | N\u002FA |\r\n| Handle iframes     | ✅ | ✅ | ❌ |\r\n| Open several tabs  | ✅ | ⏳ | ✅  |\r\n| Highlight elements | ✅ | ✅  | ✅ |\r\n\r\n\r\n✅ supported  \r\n⏳ coming soon  \r\n❌ not supported \r\n\r\n## 🔎 Support\r\n\r\nIf you're experiencing any issues getting started with LaVague, you can:\r\n\r\n- Check out our [troubleshooting guide](https:\u002F\u002Fdocs.lavague.ai\u002Fen\u002Flatest\u002Fdocs\u002Fget-started\u002Ftroubleshoot\u002F) where we list information and fixes for common issues.\r\n- Opening a [GitHub issue](https:\u002F\u002Fgithub.com\u002Flavague-ai\u002FLaVague\u002Fissues) describing your issue\r\n- Messaging us in the '#support channel' on our [Discord](https:\u002F\u002Fdiscord.gg\u002FSDxn9KpqX9\") server\r\n\r\n## 🙋 Contributing\r\n\r\nWe would love your help and support on our quest to build a robust and reliable Large Action Model for web automation.\r\n\r\nTo avoid having multiple people working on the same things & being unable to merge your work, we have outlined the following contribution process:\r\n\r\n1) 📢 We outline tasks using [`GitHub issues`](https:\u002F\u002Fgithub.com\u002Flavague-ai\u002FLaVague\u002Fissues): we recommend checking out issues with the [`help-wanted`](https:\u002Fgithub.com\u002Flavague-ai\u002FLaVague\u002Flabels\u002Fhelp%20wanted) & [`good first issue`](https:\u002F\u002Fgithub.com\u002Flavague-ai\u002FLaVague\u002Flabels\u002Fgood%20first%20issue) labels\r\n2) 🙋‍♀️ If you are interested in working on one of these tasks, comment on the issue! \r\n3) 🤝 We will discuss with you and assign you the task with a [`community assigned`](https:\u002F\u002Fgithub.com\u002Flavague-ai\u002FLaVague\u002Flabels\u002Fcommunity-assigned) label \r\n4) 💬 We will then be available to discuss this task with you\r\n5) ⬆️ You should submit your work as a PR\r\n6) ✅ We will review & merge your code or request changes\u002Fgive feedback\r\n\r\nPlease check out our [`contributing guide`](https:\u002F\u002Fdocs.lavague.ai\u002Fen\u002Flatest\u002Fdocs\u002Fcontributing\u002Fcontributing\u002F) for more details.\r\n\r\n## 🗺️ Roadmap\r\n\r\nTo keep up to date with our project backlog [here](https:\u002F\u002Fgithub.com\u002Forgs\u002Flavague-ai\u002Fprojects\u002F1\u002Fviews\u002F2).\r\n\r\n## 💰 How much does it cost to run an agent?\r\n\r\nLaVague uses LLMs, (by default OpenAI's `gpt4-o` but this is completely customizable), under the hood.\r\n\r\nThe cost of these LLM calls depends on: \r\n- the models chosen to run a given agent\r\n- the complexity of the objective\r\n- the website you're interacting with. \r\n\r\nPlease see our [dedicated documentation on token counting and cost estimations](https:\u002F\u002Fdocs.lavague.ai\u002Fen\u002Flatest\u002Fdocs\u002Fget-started\u002Ftoken-usage\u002F) to learn how you can track all tokens and estimate costs for running your agents.\r\n\r\n## 📈 Data collection\n\nWe want to build a dataset that can be used by the AI community to build better Large Action Models for better Web Agents. You can see our work so far on building community datasets on our [BigAction HuggingFace page](https:\u002F\u002Fhuggingface.co\u002FBigAction).\n\nThis is why LaVague collects the following user data telemetry by default:\n\n- Version of LaVague installed\n- Code \u002F List of actions generated for each web action step\n- The past actions\n- The \"observations\" (method used to check the current page)\n- LLM used (i.e GPT4)\n- Multi modal LLM used (i.e GPT4)\n- Randomly generated anonymous user ID\n- Whether you are using a CLI command (lavague-qa for example), the Gradio demo or our library directly.\n- The objective used \n- The chain of thoughts on the agent\n- The interaction zone on the page (bounding box)\n- The viewport size of your browser\n- The current step\n- The instruction(s) generated & the current engine used\n- The token costs & usages\n- The URL you performed an action on\n- Whether the action failed or succeeded\n- The extra used data specified\n- Error message, where relevant\n- The source nodes (chunks of HTML code retrieved from the web page to perform this action)\n\n**Be careful to NEVER includes personal information in your objectives and the extra user data. If you intend to includes personal information in your objectives\u002Fextra user data, it is HIGHLY recommended to turn off the telemetry.**\r\n\r\n### 🚫 Turn off all telemetry\r\n\r\nIf you want to turn off all telemetry, you should set the `LAVAGUE_TELEMETRY` environment variable to `\"NONE\"`.\r\n\r\nFor guidance on how to set your `LAVAGUE_TELEMTRY` environment variable, see our guide [here](https:\u002F\u002Fdocs.lavague.ai\u002Fen\u002Flatest\u002Fdocs\u002Fget-started\u002FFAQs\u002F#how-can-i-set-environment-variables).\r\n","LaVague 是一个用于开发AI网页代理的大型动作模型框架。其核心功能包括一个世界模型，该模型根据目标和当前网页状态生成指令集，以及一个动作引擎，将这些指令编译成可执行代码（如Selenium或Playwright）并执行。此外，基于LaVauge框架还开发了专门面向质量保证工程师的工具LaVague QA，能够自动将Gherkin规格转换为易于集成的测试用例，极大地提高了网页测试效率。此项目适用于需要自动化复杂网页交互场景的应用开发，比如自动化软件安装指南生成、网页应用测试等。",2,"2026-06-11 03:40:07","high_star"]