Large language models are increasingly expected to execute complex workflows whose success depends on maintaining interdependent constraints and producing artifacts that satisfy strict end-to-end veri...
LongWoF-Bench is a new benchmark of 778 machine‑verifiable long‑workflow tasks spanning code generation, agent‑environment synthesis, mathematical reasoning, and rule following. The study shows that EvoMap Genes—structured representations of verifier‑confirmed execution trajectories—outperform the Skill baseline by 8.7–15.5 percentage points across seven models, and for Claude Opus they enable 39 additional task completions while cutting token consumption by 9.9%. The results demonstrate that verified execution experience can be externalized and reused, improving long‑workflow completion without repeatedly discovering new strategies.
Evolver is an experience-driven framework that converts agent exploration into persistent, reusable capabilities without updating foundation model parameters. It uses structured Gene assets to acquire, retrieve, and evolve experience, enabling experience-conditioned execution across existing agent harnesses. In evaluations with Claude Code and Codex on several benchmarks, Evolver increased task success from 83.13% to 87.50% and reduced token consumption by up to 24.24%, showing both effectiveness and efficiency gains.
arXiv:2610.35616v1 Announce Type: new Abstract: Skills provide agents with reusable experience to guide future decisions without modifying the underlying model. However, coding trajectories are often...
arXiv:2606. 01286v1 Announce Type: cross Abstract: The rapid progress of frontier large language models has led to widespread benchmark saturation, limiting the ability of existing datasets to differentiate model capabilities or provide useful training signal.
The paper introduces an online skill‑evolution framework that transforms interaction traces and evaluator feedback into a persistent, versioned library of reusable procedures for computer‑use agents. By executing each iteration against a frozen library snapshot, the system updates skills without altering the underlying model parameters. Experiments across four OSWorld domains show that the evolving library consistently outperforms an empty‑library baseline, with gains ranging from 5.7 to 18.6 percentage points, while also revealing domain‑specific temporal stability and challenges in skill retrieval and revision.
arXiv:2608. 16544v1 Announce Type: cross Abstract: Agents increasingly rely on reusable skills to encode task knowledge, tool-use procedures, and validation rules.
arXiv:2610.28386v1 Announce Type: cross Abstract: Visual tool use enables vision-language models (VLMs) to acquire task-relevant evidence, but its effectiveness can be limited by predefined visual op...
SkillNEO introduces a non‑elitist evolutionary framework for refining agent skills, allowing lower‑reward versions to receive further edits and preserve useful partial repairs. By separating the selection of a version to use from the selection of versions to improve, the method identifies which steps to keep and which failures to fix, even revisiting earlier versions with new evidence. Experiments on 86 SkillsBench tasks across eight domains show significant reward gains for GPT‑5.4 Mini and substantial improvements when the evolved skills are reused by stronger models such as GPT‑5.5 and GPT‑5.6 Luna.
arXiv:2608.28638v1 Announce Type: new Abstract: Agent skills are portable packages of instructions and resources an agent consults at deployment. Self-evolving them fails in two ways today. First, sk...
Professional graphic design is a long-horizon agentic task in which structured, editable artifacts emerge from many interdependent actions, yet outcomes admit no reliable programmatic oracle. We intro...
arXiv:2608. 02636v1 Announce Type: cross Abstract: Self-evolving skill systems promise to improve agents by turning execution feedback into persistent skill updates without changing the underlying model.