SkillReducer: Optimizing LLM Agent Skills for Token Efficiency

• URL: http://arxiv.org/abs/2603.29919v1
• Date: Tue, 31 Mar 2026 15:57:53 GMT
• Title: SkillReducer: Optimizing LLM Agent Skills for Token Efficiency
• Abstract summary: LLM-based coding agents rely on emphskills, pre-packaged instruction sets that extend agent capabilities.<n>Every token of skill content injected into the context window incurs both monetary cost and attention dilution.<n>We present textscSkillReducer, a two-stage optimization framework.
• Score: 19.2748735514124
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: LLM-based coding agents rely on \emph{skills}, pre-packaged instruction sets that extend agent capabilities, yet every token of skill content injected into the context window incurs both monetary cost and attention dilution. To understand the severity of this problem, we conduct a large-scale empirical study of 55,315 publicly available skills and find systemic inefficiencies: 26.4\% lack routing descriptions entirely, over 60\% of body content is non-actionable, and reference files can inject tens of thousands of tokens per invocation. Motivated by these findings, we present \textsc{SkillReducer}, a two-stage optimization framework. Stage~1 optimizes the routing layer by compressing verbose descriptions and generating missing ones via adversarial delta debugging. Stage~2 restructures skill bodies through taxonomy-driven classification and progressive disclosure, separating actionable core rules from supplementary content loaded on demand, validated by faithfulness checks and a self-correcting feedback loop. Evaluated on 600 skills and the SkillsBench benchmark, \textsc{SkillReducer} achieves 48\% description compression and 39\% body compression while improving functional quality by 2.8\%, revealing a \emph{less-is-more} effect where removing non-essential content reduces distraction in the context window. These benefits transfer across five models from four families with a mean retention of 0.965, and generalize to an independent agent framework.

Related papers

• SkillsMetric: Mapping the Detection Boundary of Static Analysis for Malicious Agent Skills [12.451393173022344]
  We present textscSkillsMetric, a five-stage static analysis framework that scores skill packages along pattern density, statistical anomaly, dataflow taint, import anomaly, and capability mismatch dimensions.<n>We construct an adversarial evaluation dataset of 2266 skills spanning 16attack types across code-level, system-level, and semantic-level threats, and evaluate on the full SkillMD-138K corpus.<n>Our framework achieves an AUC of 0.93 and 5-fold cross-validated F1 of 73.4%$pm$0.5%, with strong detection of data ex
  arXiv  Detail & Related papers  (2026-08-09T04:19:10Z)
• Fantastic Adaptive Taxonomies and How to Use Them [100.20614173157708]
  AdaMAST converts execution traces into a compact, evidence-grounded failure taxonomy.<n>No code is hand-authored, no trace human-annotated.<n>In agent-system search, taxonomy-coded diagnoses of failed candidates outperform free-form.
  arXiv  Detail & Related papers  (2026-07-17T17:41:16Z)
• SkillDAG: Self-Evolving Typed Skill Graphs for LLM Skill Selection at Scale [54.70985426016736]
  We present SkillDAG, which models inter-skill relationships as a typed directed graph.<n>Each search returns vector matches, typed-edge neighbors, and conflict signals, and a propose-then-commit protocol lets the agent register execution-backed edges.<n>On ALFWorld and SkillsBench with MiniMax-M2.7, SkillDAG reaches 67.1% success and 27.3% reward.
  arXiv  Detail & Related papers  (2026-06-02T02:45:21Z)
• MOCHA: Multi-Objective Chebyshev Annealing for Agent Skill Optimization [35.57702110117083]
  MOCHA (Multi-language Chebyshev Annealing) covers the full front, including non-resident regions.<n> MOCHA breaks through on every task, achieving 7.5% relative improvement in mean correctness over the strongest baseline.
  arXiv  Detail & Related papers  (2026-05-19T04:07:41Z)
• LongSeeker: Elastic Context Orchestration for Long-Horizon Search Agents [63.75550089045995]
  Long-horizon search agents must manage a rapidly growing working context as they reason, call tools, and observe information.<n>We propose that effective context management should be adaptive: parts of the agent's trajectory are maintained at different levels of detail depending on their current relevance to the task.<n>We develop LongSeeker, a long-horizon search agent fine-tuned from Qwen3-30B-A3B on 10k synthesized trajectories.
  arXiv  Detail & Related papers  (2026-05-06T17:54:16Z)
• The Structural Attention Tax: How Retrieval Format Hijacks In-Context Learning Independent of Content [11.226669792850645]
  We formalise a phenomenon we term the structural attention tax: knowledge graph (KG) triples.<n>KG triples capture 2-3x more attention per token than semantically equivalent natural-language text.<n>We develop a formal framework decomposing attention scores into semantic and structural components.
  arXiv  Detail & Related papers  (2026-04-21T08:47:00Z)
• Experience Compression Spectrum: Unifying Memory, Skills, and Rules in LLM Agents [9.989306175511238]
  LLM agents scale to long-horizon, multi-session deployments, efficiently managing accumulated experience becomes a critical bottleneck.<n>We propose the emphExperience Compression Spectrum, a unifying framework that positions memory, skills, and rules as points along a single axis of increasing compression.
  arXiv  Detail & Related papers  (2026-04-17T09:26:25Z)
• APEX-EM: Non-Parametric Online Learning for Autonomous Agents via Structured Procedural-Episodic Experience Replay [7.370176470430802]
  We present textbfAPEX-EM, a non-parametric online learning framework that accumulates, retrieves, and reuses structured procedural plans.<n>APEX-EM achieves 89.6% accuracy versus 41.3% without memory (+48.3pp), surpassing the oracle-retrieval upper bound.<n>On BigCodeBench, it reaches 83.3% SR from a 53.9% baseline (+29.4pp), exceeding MemRL'scitememrl2025 +11.0pp gain under comparable frozen-backbone conditions.
  arXiv  Detail & Related papers  (2026-03-31T00:24:56Z)
• SkillRouter: Skill Routing for LLM Agents at Scale [18.540967600258607]
  Given a user task, the system must identify relevant skills before downstream planning or execution.<n>Existing agent stacks often rely on progressive disclosure, exposing only skill names and descriptions while hiding the full implementation body.<n>We present Skill, a compact 1.2B full-text retrieve-and-rerank pipeline.
  arXiv  Detail & Related papers  (2026-03-23T18:23:59Z)
• Prune Redundancy, Preserve Essence: Vision Token Compression in VLMs via Synergistic Importance-Diversity [54.95089105944234]
  Vision-language models (VLMs) face significant computational inefficiencies caused by excessive generation of visual tokens.<n>We propose PruneSID, a training-free Synergistic Importance-Diversity approach featuring a two-stage pipeline.<n>We demonstrate state-of-the-art performance, achieving 96.3% accuracy on LLaVA-1.5 with only 11.1% token retention, and 92.8% accuracy at extreme compression rates (5.6%) on LLaVA-NeXT.
  arXiv  Detail & Related papers  (2026-03-10T10:31:58Z)
• Gated Sparse Attention: Combining Computational Efficiency with Training Stability for Long-Context Language Models [0.0]
  Gated Sparse Attention (GSA) is an architecture that realizes the benefits of both sparse and gated attention.<n>GSA incorporates a gated lightning indexer with sigmoid activations that produce bounded, interpretable selection scores.
  arXiv  Detail & Related papers  (2026-01-12T20:33:39Z)
• D2Pruner: Debiased Importance and Structural Diversity for MLLM Token Pruning [49.16227597771663]
  D2Pruner is a framework that combines debiased importance with a structural pruning mechanism.<n>It reduces FLOPs by 74.2% while retaining 99.2% of its original performance.<n>It marks a significant advancement with up to 63. 53% improvement over existing methods.
  arXiv  Detail & Related papers  (2025-12-22T14:42:31Z)
• CIMemories: A Compositional Benchmark for Contextual Integrity of Persistent Memory in LLMs [62.116710797795314]
  Large Language Models (LLMs) increasingly use persistent memory from past interactions to enhance personalization and task performance.<n>We present CIMemories, a benchmark for evaluating whether LLMs appropriately control information flow from memory based on task context.
  arXiv  Detail & Related papers  (2025-11-18T21:51:23Z)
• TeaRAG: A Token-Efficient Agentic Retrieval-Augmented Generation Framework [62.66056331998838]
  TeaRAG is a token-efficient agentic RAG framework capable of compressing both retrieval content and reasoning steps.<n>Our reward function evaluates the knowledge sufficiency by a knowledge matching mechanism, while penalizing excessive reasoning steps.
  arXiv  Detail & Related papers  (2025-11-07T16:08:34Z)
• Eigen-1: Adaptive Multi-Agent Refinement with Monitor-Based RAG for Scientific Reasoning [53.45095336430027]
  We develop a unified framework that combines implicit retrieval and structured collaboration.<n>On Humanity's Last Exam (HLE) Bio/Chem Gold, our framework achieves 48.3% accuracy.<n>Results on SuperGPQA and TRQA confirm robustness across domains.
  arXiv  Detail & Related papers  (2025-09-25T14:05:55Z)

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.