browser-search
An open-source skill for AI agents that helps developers run self-hosted web search, page browsing, and research-material collection, rather than serving as a general browser or full search engine.
Tool overview
Adoption view: browser-search is best understood as an open-source web access layer for AI agents, not a standalone search engine and not a consumer web browser. The current evidence is mainly the official GitHub repository text, which supports its positioning, listed components, and self-hosted design, but offers limited proof on reliability, performance ceilings, or real-world deployment outcomes.
In practice, it connects SearXNG, Camofox, and CloakBrowser so an agent can search the web, open pages, and attempt to deal with some protected access scenarios. A more accurate analogy is an agent web-research skill or browsing adapter inside a research workflow. The “anti-hallucination by design” claim is a project statement; the evidence supports that it enables online retrieval, but not that final answer quality is consistently superior.
On cost and setup, the repo describes it as self-hosted, free, and unlimited. That should be treated as official project messaging about software availability, not as a guarantee of zero operating cost. Real cost likely depends on hosting SearXNG, browser infrastructure, and maintenance.
This tool does not have related social references to display yet.