#1019 · Primary category: AI Agents & Automation

browserpilot

browser-automation generative-ai gpt-3 robotic-process-automation rpa selenium selenium-python

Natural language browser automation

Project last updated:12/21/24

GitHub Stars

635

Forks

75

Contributors

6

License

MIT

Why we included this project

BrowserPilot takes a different route from the usual Selenium grind: instead of hand-writing selectors that break the moment a site changes its markup, you describe the steps in plain English and a language model compiles them into browser actions. A task like searching a site and clicking the first matching result becomes a short readable prompt rather than a pile of locators. It is honest about the catch, though. Prompts work best when you phrase them in Selenium's vocabulary, so it helps to know how the underlying tool thinks, and you can cache compiled instructions to avoid repeated API calls. Reusable functions and a memory index for visited pages push it past a one-shot script generator. For developers and researchers who want to prototype browser automation or scraping quickly, it is a useful reference point for how far natural-language control of the browser can go.

Articles for this project

No articles for this project yet.

To suggest a topic or contribute an article, contact us.

Related projects in this category