SkillGen: Verified Inference-Time Agent Skill Synthesis
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866918494856019968 |
|---|---|
| author | Ma, Yuchen Huang, Yue Bao, Han Zhuang, Haomin Shukla, Swadheen Galley, Michel Zhang, Xiangliang Feuerriegel, Stefan |
| author_facet | Ma, Yuchen Huang, Yue Bao, Han Zhuang, Haomin Shukla, Swadheen Galley, Michel Zhang, Xiangliang Feuerriegel, Stefan |
| contents | Skills are a promising way to improve LLM agent capabilities without retraining, while keeping the added procedure reusable and controllable. However, high-quality skills are still largely written by hand. We introduce SkillGen, a multi-agent framework that synthesizes a single auditable skill from trajectories generated by a base agent. The output is a human-readable artifact that can be inspected before use. Rather than merely summarizing trajectories, SkillGen leverages contrastive induction over both successful and failed trajectories to identify reusable success patterns, recurring failure modes, and behaviors that appear in nearby successes but are missing from failures. SkillGen then generates candidate skills and iteratively refines the skill. A key novelty in SkillGen is that we model agent skills as interventions to empirically verify the net effect of skills on the overall performance. Specifically, we compare outcomes on the same instances with and without the skill, so that we account for both repairs (cases where the skill fixes a baseline failure) and regressions (cases where the skill breaks a baseline success). Across a broad range of agents and datasets, SkillGen consistently improves held-out performance, outperforms existing skill-generation baselines, and produces skills that transfer across models. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2605_10999 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | SkillGen: Verified Inference-Time Agent Skill Synthesis Ma, Yuchen Huang, Yue Bao, Han Zhuang, Haomin Shukla, Swadheen Galley, Michel Zhang, Xiangliang Feuerriegel, Stefan Machine Learning Artificial Intelligence Multiagent Systems Skills are a promising way to improve LLM agent capabilities without retraining, while keeping the added procedure reusable and controllable. However, high-quality skills are still largely written by hand. We introduce SkillGen, a multi-agent framework that synthesizes a single auditable skill from trajectories generated by a base agent. The output is a human-readable artifact that can be inspected before use. Rather than merely summarizing trajectories, SkillGen leverages contrastive induction over both successful and failed trajectories to identify reusable success patterns, recurring failure modes, and behaviors that appear in nearby successes but are missing from failures. SkillGen then generates candidate skills and iteratively refines the skill. A key novelty in SkillGen is that we model agent skills as interventions to empirically verify the net effect of skills on the overall performance. Specifically, we compare outcomes on the same instances with and without the skill, so that we account for both repairs (cases where the skill fixes a baseline failure) and regressions (cases where the skill breaks a baseline success). Across a broad range of agents and datasets, SkillGen consistently improves held-out performance, outperforms existing skill-generation baselines, and produces skills that transfer across models. |
| title | SkillGen: Verified Inference-Time Agent Skill Synthesis |
| topic | Machine Learning Artificial Intelligence Multiagent Systems |
| url | https://arxiv.org/abs/2605.10999 |