Metraipolis › archive › dc_f38e9ea3dfac

ToolSpace: Scaling Tool-Centric Environment Synthesis for Long-Horizon Agents — arXiv:2611.05036

Source
arXiv (en)
Original
https://arxiv.org/abs/2611.05036
License (as archived)
CC0-1.0
Archived
2026-11-04T18:38:47.126Z
SHA-256 of the archived text
b5f9c5ab1f418fce824912004c05864944ced3c5af6452e14c097488cd2b103c
Stable URL of this page
https://metraipolis.com/documento/dc_f38e9ea3dfac

How to cite

ToolSpace: Scaling Tool-Centric Environment Synthesis for Long-Horizon Agents — arXiv:2611.05036 — arXiv (en), https://arxiv.org/abs/2611.05036 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_f38e9ea3dfac (Metraipolis archive dc_f38e9ea3dfac, sha256 b5f9c5ab1f418fce824912004c05864944ced3c5af6452e14c097488cd2b103c)
BibTeX
@misc{dcf38e9ea3dfac,
  title = {ToolSpace: Scaling Tool-Centric Environment Synthesis for Long-Horizon Agents — arXiv:2611.05036},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2611.05036},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_f38e9ea3dfac; sha256 b5f9c5ab1f418fce824912004c05864944ced3c5af6452e14c097488cd2b103c},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.05036. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

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.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.CL, nuovi lavori del giorno).
Categorie arXiv: Computation and Language (cs.CL).
Pagina del lavoro: https://arxiv.org/abs/2611.05036 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-11-04T18:32:52.169Z dalla pagina https://arxiv.org/list/cs.CL/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).