Gecko: A Simulation Environment with Stateful Feedback for Refining Agent Tool Calls

TL;DR

Gecko is a simulation environment that provides stateful feedback to improve large language model (LLM) tool calls, reducing errors and unsafe refinements by simulating tool responses and evaluating task completion.

Contribution

We introduce Gecko, a simulation environment that enhances LLM tool call accuracy through rule-based and LLM-generated feedback, enabling safer and more effective iterative refinement.

Findings

GATS improves tool calling performance across multiple LLMs

Gecko reduces errors in tool call inputs and outputs

The environment enhances task success rates in benchmark tests

Abstract

The ability to use tools is fundamental for large language model (LLM) agents. Given a task, existing systems use LLMs to plan and generate tool calls, which are executed by real-world tools to complete the task. However, tool calls are prone to errors because they are derived merely from LLM intrinsic capabilities. What is more, while it is useful to let LLMs iteratively refine the tool-call sequence using execution results from real tools, this process can be expensive and lead to unsafe results. To improve LLM tool calls and address issues caused by using real tools for refinement, we introduce Gecko, a comprehensive environment that simulates tool responses using a combination of rules and LLMs. Specifically, Gecko checks the validity of tool calls including input arguments and tool names, synthesizes reasonable responses that adhere to the output schema, and assesses whether all task objectives have been achieved. These three types of feedback provided by Gecko allow LLMs to refine their tool calls, forming a simple yet effective test-time scaling method named GATS. On BFCLv3 and $\tau^2$-bench, GATS consistently improves the tool calling performance of various LLMs including GPT-4o, GPT-5, and Gemini-3.0-pro. We further discuss working mechanisms of our method and share future possibilities.