browser

name	browser
description	Control a real, local web browser to search, navigate, and extract information.
dependencies	{"python":[],"node":[],"binaries":[]}

Web Browser 🌐

LimeBot's window to the live internet. It uses a local instance of Chrome/Chromium to interact with websites just like a human.

Core Commands:

• web_search(query, count, kind): Preferred web search. kind='news' for recent news. Returns ranked titles/URLs/snippets.
• image_search(query): Find images. Returns Image URLs + source pages.
• deep_research(query): Multi-source research with a cited synthesized answer. Use for questions needing several sources.
• browser_navigate(url): Open a page. Returns the page title and a list of interactive elements with IDs (e.g., [e12]).
• browser_click(element_id): Interact with buttons or links using the IDs from the navigation result.
• browser_type(element_id, text): Fill out forms and search bars.
• browser_scroll(direction='down'): Move through a page to reveal more content.
• browser_extract(selector='body'): Get the text content of a page.
• google_search(query): Legacy alias for web_search.

Strategy:

1. Search (web_search / image_search / deep_research) or Navigate.
2. Snapshot to see the elements.
3. Click or Type to interact.
4. Extract the final information needed.

Sending a picture to the user:

When the user asks you to send/show a picture of something, call image_search(query=...), pick the best result, then call send_media(path='<Image URL>'). send_media downloads the remote URL and delivers it as a real image in web, Discord, and WhatsApp — do not just paste the raw URL as text.