browserbase/stagehand @browserbasehq/stagehand@3.5.0 on GitHub

#2149 6e75725 Thanks @miguelg719! - Added a screenshot option to extract() that sends the current viewport screenshot with the a11y tree for extraction.
#2127 78bcde8 Thanks @seanmcguire12! - Add ignoreDefaultArgs option to selectively remove chrome-launcher's built-in default flags (e.g. --disable-extensions) when running locally
#2160 49575d6 Thanks @monadoid! - Forward constructor and request model configuration when initializing API-backed sessions.
#2171 dc1445d Thanks @seanmcguire12! - add native clipboard API

#2146 3a53ed4 Thanks @shriyatheunicorn! - Pass local browser launch options through when attaching over CDP so explicit viewport settings are respected.
#2107 8fc16d2 Thanks @miguelg719! - Fix Anthropic CUA triple_click action mapping.
#2118 3e95a87 Thanks @monadoid! - Add Vertex auth parameters to the core and server API schemas.
#2126 ebbdcd3 Thanks @seanmcguire12! - fix(core): import ToolSet from ai public export
#2120 12703a6 Thanks @miguelg719! - Fix structuredOutputMode for newer Anthropic models
#2170 1db5f1c Thanks @Kylejeong2! - Add Claude Opus 4.8 to the supported CUA model whitelist.
#2116 cb586a1 Thanks @seanmcguire12! - include "[selected]" or "[checked]" state in snapshot
#2129 765861c Thanks @miguelg719! - Add a backend-selectable v3 evaluator facade while preserving the legacy evaluator path.
#2157 2cd60a3 Thanks @miguelg719! - Add verifier trajectory, rubric, and evaluation-result types with normalized public naming.
#2131 e102a89 Thanks @miguelg719! - Capture verifier trajectory evidence from agent evidence callbacks for offline scoring.