AgentRob: From Virtual Forum Agents to Hijacked Physical Robots

TL;DR

AgentRob introduces a novel framework enabling autonomous virtual forum agents to interpret online commands and control physical robots, bridging online community interactions with real-world robotic actions.

Contribution

This work presents the first system that integrates LLM-powered forum agents with physical robots through the Model Context Protocol, enabling multi-agent coordination in online forums.

Findings

Multiple concurrent agents with distinct identities can co-exist in forums.

Forum-mediated commands can successfully control physical robots.

The framework demonstrates feasible multi-agent robot orchestration via online interactions.

Abstract

Large Language Model (LLM)-powered autonomous agents have demonstrated significant capabilities in virtual environments, yet their integration with the physical world remains narrowly confined to direct control interfaces. We present AgentRob, a framework that bridges online community forums, LLM-powered agents, and physical robots through the Model Context Protocol (MCP). AgentRob enables a novel paradigm where autonomous agents participate in online forums--reading posts, extracting natural language commands, dispatching physical robot actions, and reporting results back to the community. The system comprises three layers: a Forum Layer providing asynchronous, persistent, multi-agent interaction; an Agent Layer with forum agents that poll for @mention-targeted commands; and a Robot Layer with VLM-driven controllers and Unitree Go2/G1 hardware that translate commands into robot primitives via iterative tool calling. The framework supports multiple concurrent agents with distinct identities and physical embodiments coexisting in the same forum, establishing the feasibility of forum-mediated multi-agent robot orchestration.