# HTAA: Enhancing LLM Planning via Hybrid Toolset Agentization & Adaptation

Source: [HTAA: Enhancing LLM Planning via Hybrid Toolset Agentization & Adaptation](<https://www.sota2.com/research/paper/69ddca637d0172140ec79dff>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-04-15T05:00:11.244Z

Authors: Chengrui Huang, Junshuo Zhang, Zhiyuan Ma, Xikun Wang, Ximeng Wang, Menghua Jiang, Gang Zeng, Zhaobing Han, Shen Gao, Shuo Shang

Published: 2026-04-13T02:33:50.000Z

## Abstract

Enabling large language models to scale and reliably use hundreds of tools is critical for real-world applications, yet challenging due to the inefficiency and error accumulation inherent in flat tool-calling architectures. To address this, we propose Hybrid Toolset Agentization & Adaptation (HTAA), a hierarchical framework for scalable tool-use planning. We propose a novel toolset agentization paradigm, which encapsulates frequently co-used tools into specialized agent tools, thereby reducing the planner's action space and mitigating redundancy. To ensure effective coordination, we design Asymmetric Planner Adaptation, a trajectory-based training paradigm that aligns the high-level planner with agent tools via backward reconstruction and forward refinement. To validate the performance of HTAA, we conduct experiments on a real-world internal dataset, InfoVerify, based on the POI validation workflow of China's largest online large-scale ride-hailing platform, featuring long-horizon executable tool trajectories. Experiments on InfoVerify and widely-used benchmarks show that HTAA consistently achieves higher task success rates, requires short tool calling trajectories, and significantly reduces context overhead compared to strong baselines. Furthermore, in a production deployment, HTAA substantially reduces manual validation effort and operational cost, demonstrating its practical efficacy.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

- [Information Verification on InfoVerify proposed (real-world)](<https://www.sota2.com/research/sota/information-verification-on-infoverify-proposed-real-world.md>): Information Verification; InfoVerify proposed (real-world); Error Rate: 9.8.

- [Tool Use on Extended BFCL Single](<https://www.sota2.com/research/sota/tool-use-on-extended-bfcl-single.md>): Tool Use; Extended BFCL Single; File Accuracy: 61.54.

- [POI Information Verification on Industrial POI Verification Tasks](<https://www.sota2.com/research/sota/poi-information-verification-on-industrial-poi-verification-tasks.md>): POI Information Verification; Industrial POI Verification Tasks; AR: 84.5.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2604.10917>)

[PDF](<https://arxiv.org/pdf/2604.10917.pdf>)
