Metraipolis › archive › dc_ebc19791441e
Toollery: Scaling LLM Agents to Thousands of Skills and Tools — arXiv:2609.22218
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.22218
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-09T06:50:25.669Z
- SHA-256 of the archived text
7cbd0c6afa3968d895c618f84785ae5a38d9af7adf3cd53e7a6ebb40074d5417- Stable URL of this page
https://metraipolis.com/documento/dc_ebc19791441e
How to cite
Toollery: Scaling LLM Agents to Thousands of Skills and Tools — arXiv:2609.22218 — arXiv (en), https://arxiv.org/abs/2609.22218 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_ebc19791441e (Metraipolis archive dc_ebc19791441e, sha256 7cbd0c6afa3968d895c618f84785ae5a38d9af7adf3cd53e7a6ebb40074d5417)
BibTeX
@misc{dcebc19791441e,
title = {Toollery: Scaling LLM Agents to Thousands of Skills and Tools — arXiv:2609.22218},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.22218},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_ebc19791441e; sha256 7cbd0c6afa3968d895c618f84785ae5a38d9af7adf3cd53e7a6ebb40074d5417},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.22218. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
As LLM agents are exposed to hundreds to tens of thousands of skills, tools, and API functions, full-library prompting becomes costly, slow, and less reliable: each added candidate increases prompt tokens and latency, while longer candidate lists introduce more distractors for LLM selection. We present \textbf{Toollery}, a training-free candidate-compression framework for scalable LLM skill/tool selection. Following established document-side query expansion, Toollery generates user-intent queries from each skill/tool specification and builds a retrieval index that maps real user requests to compact candidate sets before final LLM decision-making. By treating high-level skills and atomic tools as selectable capabilities, Toollery can be applied to both skill libraries and tool registries. We evaluate Toollery on the roughly 79K-capability SkillRouter benchmark, BFCL-V4 with over 440 atomic tools, and 3,396 proprietary smart-cockpit requests over 220 tools. Across these settings, Toollery keeps online selection bounded to a compact top-$k$ candidate set and improves recall over ordinary specification retrieval. At a fixed top-10 budget, Toollery improves end-to-end selection on the cockpit dataset, and maintains comparable AST Accuracy on BFCL-V4. These results support Toollery as a practical candidate-compression framework for large and evolving agent capability libraries, while showing that quality and cost gains depend on workload coverage and provider caching.
Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-22 in cs.CL).
Categorie arXiv: Machine Learning (cs.LG) ; Artificial Intelligence (cs.AI); Computation and Language (cs.CL).
Pagina del lavoro: https://arxiv.org/abs/2609.22218 — 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 2026-10-09T06:48:45.096Z dalla pagina https://arxiv.org/catchup/cs.CL/2026-09-22?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).