Metraipolis › archive › dc_22dce0899b37
SkillGym: Training Skill-Use Agents with Automatic Verifiable Environment Generation — arXiv:2609.37539
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.37539
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-07T16:23:56.207Z
- SHA-256 of the archived text
6e81d5198dba16ef2fb80edb15b87a09d260d61cb29f27554448c95255067f0a- Stable URL of this page
https://metraipolis.com/documento/dc_22dce0899b37
How to cite
SkillGym: Training Skill-Use Agents with Automatic Verifiable Environment Generation — arXiv:2609.37539 — arXiv (en), https://arxiv.org/abs/2609.37539 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_22dce0899b37 (Metraipolis archive dc_22dce0899b37, sha256 6e81d5198dba16ef2fb80edb15b87a09d260d61cb29f27554448c95255067f0a)
BibTeX
@misc{dc22dce0899b37,
title = {SkillGym: Training Skill-Use Agents with Automatic Verifiable Environment Generation — arXiv:2609.37539},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.37539},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_22dce0899b37; sha256 6e81d5198dba16ef2fb80edb15b87a09d260d61cb29f27554448c95255067f0a},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.37539. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Skills equip LLM agents with professional knowledge and guidance to complete long-horizon and complex tasks. Although skills have been widely adopted in recent agent paradigms and harnesses, how to synthesize reliable training data and how to train agents for skill use remain underexplored. In this work, we propose SkillGym, an automatic pipeline to build verifiable environments, collect trajectories, and train skill-use agents. SkillGym first crawls a large volume of skills from the internet, then keeps those whose workflows can run reproducibly offline. A builder-reviewer pipeline is used to construct difficulty-controlled tasks, spanning four task types, each with a reference solution and an executable verifier. With this pipeline, we build 6.8k environments and collect 19k verified successful trajectories for supervised finetuning. Finetuning on these trajectories improves LLMs of different families and sizes, from 2B to 122B parameters across four skill-use benchmarks; Our Qwen3.5-9B SFT model outperforms the 397B untrained model on two of them. Further analysis shows that training teaches agents to invoke skills, raising the rate of reading the relevant skill from 28% to 96%, and that the gains hold across reasoning structures, extending to task types that form a minority of the training data and to skills held out from training Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-30 in cs.AI). Categorie arXiv: Artificial Intelligence (cs.AI). Pagina del lavoro: https://arxiv.org/abs/2609.37539 — 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 2026-10-07T15:54:53.634Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-30?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).