SpecStream for Browser Use
SpecStream is the live web data engine for teams and agents. Register a SpecStream search tool on a browser-use Agent with @tools.action. The agent searches the live web from your own process and opens only the pages it needs in the browser.
What you get
- An answer with its page
- Each search returns the answer when the pages print one, with the page it was read from, and the top 10 results with title, URL, snippet and date.
- Works with allowed_domains
- Agent(web_search=True) raises with Browser(allowed_domains=...). A tool registered with @tools.action runs in your process, and the browser keeps its domain limit.
- Pages read in one call
- A second tool reads up to 20 pages as Markdown through the Extract API, so the agent needs no browser steps to read a page's text.
Set it up
- Run pip install 'browser-use==1.0.2' httpx on Python 3.11 or newer.
- Create a key on the Search API screen of your SpecStream workspace and set it as SPECSTREAM_API_KEY.
- Register specstream_search with @tools.action and pass tools=tools to the Agent.
The full setup guide has every step, with what to do if something doesn’t arrive.
Also in Agent frameworks & builders
The same live web data, in the other tools of this kind.
Live web data, in Browser Use.
Ask in your own words for companies, places, people, filings or anything else the web lists.