SkillGLoW: Procedural-Family Skill Consolidation for Self-Improving Agents on Long-Horizon Task Streams
- URL: http://arxiv.org/abs/2609.02217v1
- Date: Wed, 02 Sep 2026 07:31:18 GMT
- Title: SkillGLoW: Procedural-Family Skill Consolidation for Self-Improving Agents on Long-Horizon Task Streams
- Abstract summary: We argue the missing unit of reuse is the solving procedure shared by a cluster of related tasks, and build SkillGLoW around it.<n>Across four benchmarks (mathematical reasoning, terminal automation, software repair, and embodied control) and three models, the priors gain 17.2 points (hard) over the no-skill baseline on average.
- Score: 55.90739575113059
- License: http://creativecommons.org/licenses/by/4.0/
- Abstract: LLM agents increasingly self-improve by writing and reusing textual skills, kept either as one global document or as a flat pool of per-task entries, though most of the evidence comes from domains with structurally similar tasks. On long-horizon workloads where each task demands a different solution, the two forms fail in opposite ways: the document collapses into generic discipline, while the pool inflates and its entries stay bound to the instance that wrote them. We argue the missing unit of reuse is the solving procedure shared by a cluster of related tasks, and build SkillGLoW (Global-Local Weave) around it: the local skills a task writes from its own execution are aggregated into procedural families and compressed into de-instantiated global priors, while the instance detail they hold is regenerated per task rather than stored; a commit gate admits a prior only when real execution shows it does not degrade the deployed library. Across four benchmarks (mathematical reasoning, terminal automation, software repair, and embodied control) and three models, the priors gain 17.2 points (hard) over the no-skill baseline on average, with positive gains in all 12 continual-improvement runs, and 18.0 with local regeneration, while the library holds one prior per procedural family, 3.6x more compact than the per-task pool. Under the same protocol GLoW leads a published single-document optimizer on 15 of 21 cells. Unmodified, the library lifts success on unseen ALFWorld tasks from 73.9% to 83.9%, evidence that what transfers is procedure rather than task memory.
Related papers
- From Interaction Traces to Persistent Skills: Online Evolution for Computer-Use Agents [48.50635499112162]
We present an online skill-evolution framework that converts interaction trajectories and evaluator feedback into a persistent, versioned library of reusable procedures.<n>We compare the full evolving-library system with a configuration-matched empty-library control across four OSWorld application domains.<n>In GIMP, provenance-aware analysis reveals retrieval across task-of-origin boundaries and revision, where repeated accepted edits fail to recover the originating task.
arXiv Detail & Related papers (2026-09-04T08:29:45Z) - Tencent WorkBuddy Bench: A Multi-Domain Coding-Agent Benchmark with Contamination-Resistant Task Construction [57.138342889101345]
Tencent WorkBuddy Bench is a multi-domain evaluation suite for coding agents.<n>This report documents its construction methodology, scoring protocol, and a cross-model leaderboard.
arXiv Detail & Related papers (2026-07-23T04:34:06Z) - Do AI Agents Know When a Task Is Simple? Toward Complexity-Aware Reasoning and Execution [1.733151600403503]
Large language model (LLM) agents increasingly automate multi-step engineering and informatics.<n>They often follow a maximum-context-first strategy--re-reading files and dependencies they have already seen--turning a one-line edit into a small code-base audit.<n>We argue the missing capability is task-aware execution-scope estimation.
arXiv Detail & Related papers (2026-07-14T17:59:31Z) - Atomic Task Graph: A Unified Framework for Agentic Planning and Execution [13.978062792065215]
Atomic Task Graph (ATG) is a unified control framework for planning and execution.<n>ATG maintains an explicit graph to expose dependencies and support reuse.<n>ATG consistently outperforms strong baselines in success rate and execution efficiency.
arXiv Detail & Related papers (2026-07-02T09:34:20Z) - ClawArena-Team: Benchmarking Subagent Orchestration and Dynamic Workflows in Language-Model Agents [61.184954205350984]
ClawArena-Team is a benchmark of 41 multi-turn, multimodal, multi-agent scenarios spanning 258 evaluation rounds and 72 staged updates.<n>All scoring is execution-based with no LLM judge: an overall score -- the Subagent-Management Score (SMS) -- multiplies task correctness by a least-privilege and modality-routing factor.
arXiv Detail & Related papers (2026-06-30T06:06:08Z) - SKILL.nb: Selective Formalization and Gated Execution for Durable Agent Workflows [16.693609667845948]
We introduce SKILL.nb, a framework for governing reusable agent with evidence-versus-calibrated lifecycle policies.<n>SKILL.nb uses selective formalization: execution decides which workflow steps should become executable code.<n> gate-conditioned execution lets each step run code when its gates validate, or fall back locally when drift invalidates the executable realization.
arXiv Detail & Related papers (2026-06-06T08:27:18Z) - ClawMark: A Living-World Benchmark for Multi-Turn, Multi-Day, Multimodal Coworker Agents [77.22389710754452]
We introduce bench, a benchmark for coworker agents built around multi-turn multi-day tasks.<n>The current release contains 100 tasks across 13 professional scenarios, executed against five stateful sandboxed services.<n>The strongest model reaches 75.8 weighted score, but the best strict Task Success is only 20.0%, indicating that partial progress is common.
arXiv Detail & Related papers (2026-04-26T16:05:02Z) - Localizing Task Information for Improved Model Merging and Compression [61.16012721460561]
We show that the information required to solve each task is still preserved after merging as different tasks mostly use non-overlapping sets of weights.
We propose Consensus Merging, an algorithm that eliminates such weights and improves the general performance of existing model merging approaches.
arXiv Detail & Related papers (2024-05-13T14:54:37Z)
This list is automatically generated from the titles and abstracts of the papers in this site.
This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.