Switch to a browser window or tab by Selenium window handle.
AI agents invoke switch_window to trigger actions in SeleniumMCP. What it does depends on the arguments the agent supplies, and its effects often reach beyond the immediate call — builds kicked off, notifications sent, workflows started.
Switching browser windows/tabs is a browser automation action that changes the active context of the browser session. It's an external operation that affects the state of the browser, making it an Execute category action. Misuse could redirect browser focus unexpectedly within a session, but the blast radius is limited to the current browser session.
From the tool's definition Switch to a browser window or tab by Selenium window handle
Attacks that exploit this kind of access
Switch to a browser window or tab by Selenium window handle. It is categorised as a Execute tool in the SeleniumMCP MCP Server, which means it can trigger actions or run processes. Use rate limits and argument validation.
Register the Selenium MCP server in PolicyLayer and add a rule for switch_window: allow, deny, rate-limit, or require approval. Point your MCP client at the PolicyLayer proxy URL and the rule is enforced on every call, before it reaches SeleniumMCP. Nothing to install.
switch_window is a Execute tool with high risk. Execute tools should be rate-limited and have argument validation enabled.
Yes. Add a rate_limit block to the switch_window rule in your PolicyLayer policy. For example, setting max: 10 and window: 60 limits the tool to 10 calls per minute. Rate limits are tracked per agent session and reset automatically.
Set action: deny in the PolicyLayer policy for switch_window. The AI agent will receive a policy violation error and cannot call the tool. You can also include a reason field to explain why the tool is blocked.
switch_window is provided by the Selenium MCP server (lokii0911/seleniummcp). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Every MCP server has a record like this.
Type a name, get the same breakdown: verified identity, auth posture, risk grade, capabilities, recommended policy.
Teams ship this data inside their own products. See what a licence covers →