Comment by smpandya

Are there results comparing agents running different tools (agent-browser, playwright MCP, browse CLI), or is this mostly Stagehand focused?