In fx, the tools are registered as mcp_stagehand_run, mcp_stagehand_snapshot, and mcp_stagehand_screenshot. Select them with mcp_select_tool using those exact names before calling them; mcp_search_tools may return no results for this server, so do not rely on it. Never invent another Stagehand tool name. There is no separate stagehand_navigate, mcp_stagehand_navigate, or start tool.

You control one persistent browser through exactly three tools:

snapshot: inspect the active page and hydrate bracketed element IDs.
run: browse or navigate with JavaScript (for example, await page.goto(url)) or provide snapshot actions using the Playwright-shaped page API. The first call initializes the browser.
screenshot: inspect the rendered page visually. Prefer viewport JPEG with {"type":"jpeg","quality":40,"fullPage":false}; inline full-page PNGs may exceed fx's MCP response-frame limit.

Use snapshot actions for simple interactions and run code for multi-step workflows. Pass run exactly one of code or actions; every action uses "op" and "id", never "kind" or "ref". Snapshot IDs are valid only for the latest snapshot of the active page; snapshot again after navigation or stale IDs. Do not launch another browser.

stagehand-facade