mlsanityData to 14 Oct 2026

Bridging Thought and Action: Taming Long-Horizon Instability in Open-Source LLM Agents with a MetaTool-Enhanced ROS Framework

arXiv 2609.13335 ↗First posted 11 Sep 2026cs.RO cs.AIcited by 0

On the map: LLM reasoning → Spatial and embodied reasoning

also in 1 more
Abstract (arXiv)

Large Language Models (LLMs) have enabled more natural human-robot interaction, but open-source models often exhibit unstable long-horizon reasoning and inefficient action execution when deployed in agentic robotic frameworks. This paper presents an enhanced ROS-Agent based architecture that improves task reliability and execution efficiency for agentic robotic systems using open-source LLMs. The proposed system introduces a novel intermediate mechanism, termed the MetaTool, which enforces structured planning prior to action execution. Given a natural-language command, the MetaTool induces the LLM to generate a pseudo-code plan of intended tool invocations, which is stored in the ROS-Agent's scratchpad and persists throughout execution. By explicitly separating planning from execution, the proposed approach reduces execution loops and improves deterministic behavior. The architecture is validated on a custom mobile robotic platform with multimodal perception and motion control capabilities. Experimental results on real-world interactive tasks demonstrate improved task completion and contextual consistency, with up to ~24% gains on complex tasks compared to the baseline framework.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in Spatial and embodied reasoning, one of 17 directions in LLM reasoning. Builds on 24 of its 34 references that are in the map. First posted on arXiv 11 Sep 2026.

Builds on

24 of 34 references.

Closest ancestors · 1

Same direction, newest first

  1. 2024LLM3:Large Language Model-based Task and Motion Planning with Motion Failure Reasoning18 Mar

Other references · 23

  1. 2022ReAct: Synergizing Reasoning and Acting in Language Modelscited by 8,695
  2. 2024GPT-4o System Cardcited by 4,588
  3. 2023Toolformer: Language Models Can Teach Themselves to Use Toolscited by 4,182
  4. 2023Lost in the Middle: How Language Models Use Long Contextscited by 3,455
  5. 2025gpt-oss-120b & gpt-oss-20b Model Cardcited by 1,386
  6. 2022Code as Policies: Language Model Programs for Embodied Controlcited by 1,210
  7. 2022Language Models as Zero-Shot Planners: Extracting Actionable Knowledge for Embodied Agentscited by 1,035
  8. 2025Kimi K2: Open Agentic Intelligencecited by 567
  9. 2023SayPlan: Grounding Large Language Models using 3D Scene Graphs for Scalable Robot Task Planningcited by 294
  10. 2025Where LLM Agents Fail and How They can Learn From Failurescited by 133
  11. 2024Manipulate-Anything: Automating Real-World Robots using Vision-Language Modelscited by 83
  12. 2023ISR-LLM: Iterative Self-Refined Large Language Model for Long-Horizon Sequential Task Planningcited by 77
11 more
  1. 2023SayCanPay: Heuristic Planning with Large Language Models using Learnable Domain Knowledgecited by 49
  2. 2023Chat with the Environment: Interactive Multimodal Perception Using Large Language Modelscited by 43
  3. 2023CoPAL: Corrective Planning of Robot Actions with Large Language Modelscited by 40
  4. 2024A3VLM: Actionable Articulation-Aware Vision Language Modelcited by 33
  5. 2023Interactive Planning Using Large Language Models for Partially Observable Robotics Taskscited by 27
  6. 2024SPINE: Online Semantic Planning for Missions with Incomplete Natural Language Specifications in Unstructured Environmentscited by 18
  7. 2025Alignment for Efficient Tool Calling of Large Language Modelscited by 17
  8. 2024Enabling Novel Mission Operations and Interactions with ROSA: The Robot Operating System Agentcited by 16
  9. 2024BTGenBot: Behavior Tree Generation for Robotic Tasks with Lightweight LLMscited by 14
  10. 2024Large Language Models for Orchestrating Bimanual Robotscited by 10
  11. 2025Demonstrating the Octopi-1.5 Visual-Tactile-Language Modelcited by 9

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2024ROS-LLM: A ROS framework for embodied AI with task feedback and structured reasoning
  2. 2024LLaMAR: Long-Horizon Planning for Multi-Agent Robots in Partially Observable Environments
  3. 2026Agentic LLM Planning via Step-Wise PDDL Simulation: An Empirical Characterisation
  4. 2025Think, Act, Learn: A Framework for Autonomous Robotic Agents using Closed-Loop Large Language Models
  5. 2024ConceptAgent: LLM-Driven Precondition Grounding and Tree Search for Robust Task Planning and Execution
  6. 2026ROSClaw: An OpenClaw ROS 2 Framework for Agentic Robot Control and Interaction
  7. 2026From Language to Action: Can LLM-Based Agents Be Used for Embodied Robot Cognition?
  8. 2024MALMM: Multi-Agent Large Language Models for Zero-Shot Robotics Manipulation