Demonstration-Free Robotic Control via LLM Agents

TL;DR

This paper demonstrates that large language model (LLM) agents can control robotic manipulation tasks without demonstrations or fine-tuning, achieving high success rates across multiple benchmarks by reasoning through strategies.

Contribution

It introduces FAEA, a novel approach applying unmodified LLM agent frameworks to embodied manipulation, enabling demonstration-free control and planning in robotics.

Findings

FAEA achieves success rates of 84.9%, 85.7%, and 96% on LIBERO, ManiSkill3, and MetaWorld.

Performance approaches VLA models trained with fewer than 100 demonstrations.

Optional human feedback further improves success rate to 88.2% on LIBERO.

Abstract

Robotic manipulation has increasingly adopted vision-language-action (VLA) models, which achieve strong performance but typically require task-specific demonstrations and fine-tuning, and often generalize poorly under domain shift. We investigate whether general-purpose large language model (LLM) agent frameworks, originally developed for software engineering, can serve as an alternative control paradigm for embodied manipulation. We introduce FAEA (Frontier Agent as Embodied Agent), which applies an LLM agent framework directly to embodied manipulation without modification. Using the same iterative reasoning that enables software agents to debug code, FAEA enables embodied agents to reason through manipulation strategies. We evaluate an unmodified frontier agent, Claude Agent SDK, across the LIBERO, ManiSkill3, and MetaWorld benchmarks. With privileged environment state access, FAEA achieves success rates of 84.9%, 85.7%, and 96%, respectively. This level of task success approaches that of VLA models trained with less than 100 demonstrations per task, without requiring demonstrations or fine-tuning. With one round of human feedback as an optional optimization, performance increases to 88.2% on LIBERO. This demonstration-free capability has immediate practical value: FAEA can autonomously explore novel scenarios in simulation and generate successful trajectories for training data augmentation in embodied learning. Our results indicate that general-purpose agents are sufficient for a class of manipulation tasks dominated by deliberative, task-level planning. This opens a path for robotics systems to leverage actively maintained agent infrastructure and benefit directly from ongoing advances in frontier models. Code is available at https://github.com/robiemusketeer/faea-sim