ChainThink reports that on March 14, developer Petr Baudis stated that with the official release of Google Chrome 146, users can now enable a toggle to expose their current live browsing session to AI agents via the Model Context Protocol (MCP), allowing the agents to perform actions directly within the user’s logged-in browsing environment.
Baudis also revealed that he previously envisioned having Claude browse and perform tasks on behalf of the user while logged in, using the command: amp mcp add chrome-devtools -- npx chrome-devtools-mcp@latest --autoConnect. With the implementation of Chrome MCP capabilities, this scenario is now gradually becoming a reality.
