Fuji-Web: AI Based Full Browser Automation đź—»
Fuji-Web stands out as an intelligent AI partner designed for effortless browser automation. It understands a user's intent, navigates websites autonomously, and carries out tasks on behalf of the user, all while clearly explaining each step of the process. This makes browsing more efficient and interactive without the need for constant manual intervention.
How Does It Work?
Fuji-Web operates by integrating AI technology directly into your browser. Users can view a demo video, benchmarks, and get an in-depth technical overview by visiting this blog post. This resource provides greater insight into the mechanics behind Fuji-Web, showcasing its capabilities in real-world scenarios.
Installation and Usage
Download and Install the Extension
To get started with Fuji-Web:
- Navigate to the releases page on GitHub and download the latest "fuji-extension.zip".
- Unzip the file.
- Load the extension onto Chrome by:
- Going to
chrome://extensions/
- Enabling
Developer mode
- Selecting
Load unpacked extension
and choosing the unzipped folder
- Going to
Using the Extension
Once installed, here’s how to utilize Fuji-Web:
- Locate the Fuji icon on your browser’s toolbar and click it to open the side panel.
- You will need an OpenAI API Key or an Anthropic API key. Enter the API key in the designated box. Rest assured, your key is stored locally in your browser and is not shared with others.
- Navigate to a desired webpage and describe the task you wish Fuji-Web to execute.
Note: All prompts are directly sent to your chosen API. Fuji-Web does not collect personal information.
Building the Extension from Source
For developers interested in customization:
- Ensure you have Node.js installed.
- Clone the Fuji-Web repository.
- Globally install
pnpm
vianpm install -g pnpm
. - Execute
pnpm install
to install dependencies andpnpm dev
to start the development server orpnpm build
to create the build.
Following the build process, the extension can be loaded from the dist
folder.
Future Developments
Fuji-Web’s roadmap includes exciting enhancements such as:
- Providing an API for smooth integration with automation frameworks like Puppeteer, Playwright, and Selenium.
- Supporting more complex workflows across multiple tabs.
- Expanding capabilities to include more browsing actions and extracting content from entire pages.
- Adding features to save and share workflows, allowing users to collaborate on building knowledge bases to improve performance.
Troubleshooting and Contribution
For help, visit the Troubleshooting Guide. Fuji-Web also welcomes contributions, and interested individuals can refer to the Contribution Guide for instructions on contributing, reporting bugs, and suggesting improvements. Join the conversation on Discord at https://discord.gg/yfMjZ8udb5.
Credits
Fuji-Web draws inspiration from several sources:
- Image annotation techniques from Microsoft’s UFO paper.
- Browser extension ideas from TaxyAI's project.
- The foundation of the Chrome extension set-up was based on a project by Jonghakseo.
- The Fuji logo design is courtesy of Toss Face.
In conclusion, Fuji-Web is more than just a tool; it’s a revolutionary companion for browser automation that offers ease, efficiency, and innovation, helping users achieve their online tasks with unprecedented convenience.