ToolSpace: Scaling Tool-Centric Environment Synthesis for Long-Horizon Agents

Guanting Dong, Zhichao Sun, Xinyu Yang, Rui Wang, Xin Deng, Jiajie Jin, Zhixiong Yang, Junhao Huang, Han Han, Mingyu Ma, Xiaotao Wang, Yuyang Hu, Xiaoshuai Song, Xiaoxi Li, Jiejun Tan, Zhecan Wang, Shihan Dou, Honglin Guo, Zhiheng Xi, Tao Gui, Tao Ji, Qi Zhang, Yutao Zhu, Ji-Rong Wen, Zhicheng Dou

ToolSpace: Scaling Tool-Centric Environment Synthesis for Long-Horizon Agents: 33 upvotes on Hugging Face Daily Papers, #10 of 46 papers on 2026-11-04. Day-by-day upvote history.

Upvotes over time

It had 33 upvotes by its first day, 33 upvotes by the next day, 33 upvotes by day 3, 33 upvotes by day 7, and has 33 now. It ranked #24 of 130 papers in its week and #24 of 130 in its month. Its code repository has 28 GitHub stars.

Large language models are deployed as long-horizon agents that compose diverse tools across MCP services into executable workflows. Existing MCP environment-scaling methods are data-centric: they first synthesize a database and scenario, then derive tools and tasks from them. As a result, tool interfaces may not match real ones, tool compositions are not grounded in execution, and tasks rarely span more than one environment. We propose ToolSpace, a tool-centric framework that reverses this order. ToolSpace mines data, state, and retry dependencies from adjacent calls in execution traces of real MCP services and open agent corpora, using deterministic rules rather than LLM judgment, and clusters the resulting weighted tool-composition graph into environments. Databases are synthesized only for state-mutating environments, while query and stateless tools replay recorded returns. ToolSpace then performs state-aware random walks with rollback and decompiles each chain into a natural-language task with assertion-based checks, keeping only samples that replay successfully from the initial state. Applied to AutomationBench and tau3-Bench Banking, the pipeline produces harder Pro versions on which 7 frontier models drop by up to 38.2 points. Across seven long-horizon tool-use benchmarks, ToolSpace raises Qwen3.6-35B-A3B from 23.0 to 41.0 average score (43.2 with RL) and GLM-5.2 from 44.6 to 56.6, on par with Claude Opus 5. Ablations show that execution-witnessed edges clearly outperform LLM-judged ones, and performance keeps improving with data and environment scale.

Paper page on Hugging Face · arXiv

Data: hysts-bot-data/daily-papers-stats and the Daily Papers API. Open data: tardellirs/paper-pulse-data. Sister project: Model Pulse, the download history of every model on the Hub.