Search the web for information using the built-in `web_search` tool. Use when users ask questions requiring up-to-date information, research, or fact-checking.
Search the web using the built-in web_search tool. On Aliceloop this is the built-in research path for current information, fact-checking, and source discovery.
Before searching, check whether Chrome Relay is healthy:
ChromeRelayStatus()
web_search normally; search itself should not block on relayweb_search(query="your search query", max_results=5)
max_results small unless the task needs a wider sweepAfter finding URLs from search, use web_fetch to read the actual page content:
web_fetch(url="https://example.com/article")
If web_search or web_fetch are not enough because a page needs login, captcha, or multi-step interaction, fall back to the browser skill:
browser_navigate(url="https://example.com")
browser_snapshot()
If the page needs clicks, typing, or screenshots, continue with the native browser tools:
browser_click(ref="...")
browser_type(ref="...", text="...")
browser_screenshot()
The browser skill prefers a visible Aliceloop Desktop Chrome relay with persistent login state when available, and otherwise falls back to the local browser backend. For supported structured site adapters, the browser path may also use OpenCLI after the native browser path is ruled out.
today, latest, 今天, or 最新, use the current runtime year instead of hardcoding a fixed yearweb_fetch instead of re-searching forever"Latest AI news":
web_search(query="latest AI news <current_year>", max_results=5)
"Python 3.13 new features":
web_search(query="python 3.13 new features", max_results=5)
Fetch a specific article after search:
web_fetch(url="https://docs.python.org/3.13/whatsnew/3.13.html")