{
  "title": "VIA: target selection and local execution",
  "caption": "The author made this interface mechanism diagram. Screenshots and MCP tools let the agent select a virtual gripper waypoint. The robot backend supplies local control. The example separates observation from execution. It does not include source screenshots, measured task results or direct joint control.",
  "alt": "The visual agent reads a robot interface, sends gripper-waypoint requests with MCP tools and reads local execution results.",
  "sources": [
    "https://arxiv.org/html/2607.11119v1"
  ],
  "limitations": [
    "The execution is an example. Playback duration is not measured runtime."
  ]
}