SkillGrad: Optimizing Agent Skills Like Gradient Descent

• URL: http://arxiv.org/abs/2605.27760v1
• Date: Tue, 26 May 2026 23:18:43 GMT
• Title: SkillGrad: Optimizing Agent Skills Like Gradient Descent
• Abstract summary: Agent skills provide a lightweight way to adapt LLM agents to specialized domains by storing reusable procedural knowledge in structured files.<n>Existing skill-evolution methods often address these deficiencies through reflections without an explicit formulation.<n>We propose SkillGrad, a gradient-descent-inspired framework for optimizing agent skills.
• Score: 30.781132643932338
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Agent skills provide a lightweight way to adapt LLM agents to specialized domains by storing reusable procedural knowledge in structured files. However, whether downloaded from third parties or self-generated, these skills are often unreliable, incomplete, or outdated. Existing skill-evolution methods often address these deficiencies through heuristic reflections without an explicit optimization formulation. In this paper, we propose SkillGrad, a gradient-descent-inspired framework for optimizing agent skills. SkillGrad treats the skill package as a structured parameter to optimize in a gradient descent fashion: task executions provide trajectory-level loss evidence, automatic diagnoses then provide text-based gradients that indicate the correction directions. To stabilize optimization across iterations, a momentum agent accumulates recurring diagnostic patterns into a persistent memory overlay. Finally, an LLM-based patcher executes the parameter update by applying layer-aware edits to the skill package. Evaluated on SpreadsheetBench Verified and WikiTableQuestions, SkillGrad consistently outperforms training-based skill evolution baselines across two backbone LLMs, improving over the strongest training-based baseline by $6.7$ percentage points on average. Ablations further show that momentum and contrastive diagnosis both contribute to the final skill quality.

Related papers

• Write, Execute, Refine: From Skill Followers to Skill Optimizers via Reinforcement Learning from Execution Feedback [52.78692582507693]
  Expert-written natural language skills can improve tool-using agents, yet agent-authored skills perform 8-11 points worse than using no skill.<n>We study how to organize execution experience from intermediate skills into training states for an executor.
  arXiv  Detail & Related papers  (2026-08-18T09:52:48Z)
• SkillProx: Self-Evolving Agent Skills via Proximal Textual Gradient Descent [27.354076514186094]
  LLM agents increasingly adapt to recurring tasks by accumulating procedural knowledge in skills.<n>Recent methods refine skills through iterative task execution, failure diagnosis, and trajectory-guided text-space updates.<n>We introduce SkillProx, a proximal-gradient-inspired forward--backward framework that couples closed-loop diagnostic evolution with utility-aware proximal refinement.
  arXiv  Detail & Related papers  (2026-08-07T17:40:33Z)
• VASO: Formally Verifiable Self-Evolving Skills for Physical AI Agents [57.240036084348354]
  We introduce VASO, a framework for verification-guided self-evolution of robot skill contracts.<n>VASO verifies logically inconsistent skill contracts, then verifies plans induced by the skill against global and local temporal specifications.<n>On Clearpath Jackal and PX4 quadcopter tasks, VASO reaches 97.2% formal-specification compliance using fewer than 100 optimization samples.
  arXiv  Detail & Related papers  (2026-06-03T20:02:35Z)
• SkillRevise: Improving LLM-Authored Agent Skills via Trace-Conditioned Skill Revision [41.562357872885826]
  We propose SkillRevise, an execution-grounded framework designed to iteratively refine agent skills.<n>SkillRevise diagnoses skill defects from execution evidence, retrieves relevant repair principles from a general memory, and applies execution-anchored edits.<n>It substantially outperforms one-shot baselines, improving the base agent's success rate on SkillsBench from 36.05% to 61.63%.
  arXiv  Detail & Related papers  (2026-05-31T10:19:13Z)
• CODESKILL: Learning Self-Evolving Skills for Coding Agents [13.599750520978988]
  Coding agents produce rich trajectories while solving software-engineering tasks.<n>To enable agent self-evolution, these trajectories can be distilled into reusable procedural skills that compactly encode experience to guide future behavior.<n>We propose CODESKILL, an LLM-based framework that reformulates skill extraction and skill-bank maintenance as a learnable management policy.
  arXiv  Detail & Related papers  (2026-05-25T05:12:49Z)
• Harnessing LLM Agents with Skill Programs [58.356514745548026]
  We introduce HASP, a new framework that upgrades skills into executable Program Functions (PFs)<n>PFs act as executable guardrails that activate on failure-prone states and modify the next action or inject corrective context.<n> HASP drives substantial gains compared to both training-free and training-based methods on web-search, math reasoning, and coding tasks.
  arXiv  Detail & Related papers  (2026-05-18T01:35:11Z)
• Skill-R1: Agent Skill Evolution via Reinforcement Learning [84.35984979949502]
  Skill-R1 is a reinforcement learning framework for instance-level recurrent skill optimization from verifiable rewards.<n>It preserves black-box compatibility with both open- and closed-source models while making adaptation substantially cheaper than model-level updates.
  arXiv  Detail & Related papers  (2026-05-10T06:19:15Z)
• SkillLens: Adaptive Multi-Granularity Skill Reuse for Cost-Efficient LLM Agents [6.293920920196533]
  SkillLens is a hierarchical skill-evolution framework that organizes skills into a four-layer graph of policies, strategies, procedures, and primitives.<n>It retrieves semantically relevant skill seeds, expands them through degree-corrected random walk over the skill graph, and then uses a verifier to decide whether each visited unit should be accepted, decomposed, rewritten, or skipped.<n>Across MuLocbench and ALFWorld, SkillLens consistently improves over strong skill-based baselines.
  arXiv  Detail & Related papers  (2026-05-08T18:48:04Z)
• SkillMOO: Multi-Objective Optimization of Agent Skills for Software Engineering [9.008257157319271]
  SkillMOO improves pass rate by up to 131% while reducing cost up to 32% relative to the best baseline per task at low optimization overhead.<n>We present SkillMOO, a multi-objective optimization framework that automatically evolves skill bundles using LLM-proposed edits and NSGA-II survivor selection.<n>On three SkillsBench software engineering tasks, SkillMOO improves pass rate by up to 131% while reducing cost up to 32% relative to the best baseline per task at low optimization overhead.
  arXiv  Detail & Related papers  (2026-04-10T13:08:01Z)
• MetaClaw: Just Talk -- An Agent That Meta-Learns and Evolves in the Wild [74.7263562191605]
  Large language model (LLM) agents are increasingly used for complex tasks.<n>Existing methods either store raw trajectories without distilling knowledge, maintain static skill libraries, or require disruptive downtime for retraining.<n>We present MetaClaw, a continual meta-learning framework that jointly evolves a base LLM policy and a library of reusable behavioral skills.
  arXiv  Detail & Related papers  (2026-03-17T22:30:30Z)
• EvoSkill: Automated Skill Discovery for Multi-Agent Systems [6.319876096746374]
  We introduce textbfEvoSkill, a self-evolving framework that automatically discovers and refines agent skills.<n>EvoSkill analyzes execution failures, proposes new skills or edits to existing ones, and materializes them into structured, reusable skill folders.<n>We evaluate EvoSkill on two benchmarks: OfficeQA, a grounded reasoning benchmark over U.S. Treasury data, and SealQA, a noisy retrieval benchmark.
  arXiv  Detail & Related papers  (2026-03-03T09:07:22Z)
• GradAlign: Gradient-Aligned Data Selection for LLM Reinforcement Learning [55.03441672267886]
  We propose GradAlign, a gradient-aligned data selection method for reinforcement learning.<n>We evaluate GradAlign across three data regimes: unreliable reward signals, distribution imbalance, and low-utility training corpus.
  arXiv  Detail & Related papers  (2026-02-25T01:54:50Z)
• ET-Agent: Incentivizing Effective Tool-Integrated Reasoning Agent via Behavior Calibration [68.89572566071575]
  ETAgent is a training framework for calibrating agent's tool-use behavior.<n>It is designed to progressively calibrate erroneous behavioral patterns to optimal behaviors.
  arXiv  Detail & Related papers  (2026-01-11T11:05:26Z)

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.