# SkillRouter: Skill Routing for LLM Agents at Scale

arXiv: https://arxiv.org/abs/2603.22455 · HTML: https://mlsanity.com/paper/2603.22455/

## Key facts

- Cited by 61 papers in the map (data through 14 Oct 2026).
- Placed in Agent skill learning and evolution, one of 16 directions in LLM agents and tool use.
- Builds on 23 of its 34 references that are in the map.
- First posted on arXiv 23 Mar 2026.

## Placed in

- LLM agents and tool use / [Agent skill learning and evolution](https://mlsanity.com/agents/agent-skill-learning-and-evolution/)

## Builds on (most cited first, 23 of 23)

- [Toolformer: Language Models Can Teach Themselves to Use Tools](https://mlsanity.com/paper/2302.04761/) (2023)
- [Dense Passage Retrieval for Open-Domain Question Answering](https://mlsanity.com/paper/2004.04906/) (2020)
- [ToolLLM: Facilitating Large Language Models to Master 16000+ Real-world APIs](https://mlsanity.com/paper/2307.16789/) (2023)
- [Gorilla: Large Language Model Connected with Massive APIs](https://mlsanity.com/paper/2305.15334/) (2023)
- [Qwen3 Embedding: Advancing Text Embedding and Reranking Through Foundation Models](https://mlsanity.com/paper/2506.05176/) (2025)
- [Text Embeddings by Weakly-Supervised Contrastive Pre-training](https://mlsanity.com/paper/2212.03533/) (2022)
- [Unsupervised Dense Information Retrieval with Contrastive Learning](https://mlsanity.com/paper/2112.09118/) (2021)
- [Kimi k1.5: Scaling Reinforcement Learning with LLMs](https://mlsanity.com/paper/2501.12599/) (2025)
- [HuggingGPT: Solving AI Tasks with ChatGPT and its Friends in Hugging Face](https://mlsanity.com/paper/2303.17580/) (2023)
- [Approximate Nearest Neighbor Negative Contrastive Learning for Dense Text Retrieval](https://mlsanity.com/paper/2007.00808/) (2020)
- [Towards General Text Embeddings with Multi-stage Contrastive Learning](https://mlsanity.com/paper/2308.03281/) (2023)
- [Passage Re-ranking with BERT](https://mlsanity.com/paper/1901.04085/) (2019)
- [C-Pack: Packed Resources For General Chinese Embeddings](https://mlsanity.com/paper/2309.07597/) (2023)
- [Kimi K2.5: Visual Agentic Intelligence](https://mlsanity.com/paper/2602.02276/) (2026)
- [NV-Embed: Improved Techniques for Training LLMs as Generalist Embedding Models](https://mlsanity.com/paper/2405.17428/) (2024)
- [Is ChatGPT Good at Search? Investigating Large Language Models as Re-Ranking Agents](https://mlsanity.com/paper/2304.09542/) (2023)
- [SkillsBench: Benchmarking How Well Agent Skills Work Across Diverse Tasks](https://mlsanity.com/paper/2602.12670/) (2026)
- [MetaTool Benchmark for Large Language Models: Deciding Whether to Use Tools and Which to Use](https://mlsanity.com/paper/2310.03128/) (2023)
- [CRAFT: Customizing LLMs by Creating and Retrieving from Specialized Toolsets](https://mlsanity.com/paper/2309.17428/) (2023)
- [AnyTool: Self-Reflective, Hierarchical Agents for Large-Scale API Calls](https://mlsanity.com/paper/2402.04253/) (2024)
- [Organizing, Orchestrating, and Benchmarking Agent Skills at Ecosystem Scale](https://mlsanity.com/paper/2603.02176/) (2026)
- [Retrieval Models Aren't Tool-Savvy: Benchmarking Tool Retrieval for Large Language Models](https://mlsanity.com/paper/2503.01763/) (2025)
- [ToolRerank: Adaptive and Hierarchy-Aware Reranking for Tool Retrieval](https://mlsanity.com/paper/2403.06551/) (2024)

## Built on by (most cited first, 30 of 61)

- [SkillX: Automatically Constructing Skill Knowledge Bases for Agents](https://mlsanity.com/paper/2604.04804/) (2026)
- [How Well Do Agentic Skills Work in the Wild: Benchmarking LLM Skill Usage in Realistic Settings](https://mlsanity.com/paper/2604.04323/) (2026)
- [SkillRet: A Large-Scale Benchmark for Skill Retrieval in LLM Agents](https://mlsanity.com/paper/2605.05726/) (2026)
- [SkillsVote: Lifecycle Governance of Agent Skills from Collection, Recommendation to Evolution](https://mlsanity.com/paper/2605.18401/) (2026)
- [From Skill Text to Skill Structure: The Scheduling-Structural-Logical Representation for Agent Skills](https://mlsanity.com/paper/2604.24026/) (2026)
- [SkillMOO: Multi-Objective Optimization of Agent Skills for Software Engineering](https://mlsanity.com/paper/2604.09297/) (2026)
- [Dynamic Skill Lifecycle Management for Agentic Reinforcement Learning](https://mlsanity.com/paper/2605.10923/) (2026)
- [SkillAxe: Sharpening LLM-Authored Agent Skills Through Evaluation-Guided Self-Refinement](https://mlsanity.com/paper/2606.10546/) (2026)
- [WikiSkill: Compiling Agent Experience into Persistent Knowledge for Skill Evolution](https://mlsanity.com/paper/2608.27454/) (2026)
- [SkillRAE: Agent Skill-Based Context Compilation for Retrieval-Augmented Execution](https://mlsanity.com/paper/2605.10114/) (2026)
- [Agent Skill Evaluation and Evolution: Frameworks and Benchmarks](https://mlsanity.com/paper/2606.11435/) (2026)
- [SCOPE: Structured Decomposition and Conditional Skill Orchestration for Complex Image Generation](https://mlsanity.com/paper/2605.08043/) (2026)
- [Skill Is Not Document: Query-Conditioned Compatibility for LLM Agent Skill Routing](https://mlsanity.com/paper/2606.03565/) (2026)
- [SkillsInjector: Dynamic Skill Context Construction for LLM Agents](https://mlsanity.com/paper/2605.29794/) (2026)
- [Generative Skill Composition for LLM Agents](https://mlsanity.com/paper/2606.32025/) (2026)
- [Dynamic Agent Skills: A Lifecycle Survey and Taxonomy of Evolving Skill Libraries](https://mlsanity.com/paper/2607.10113/) (2026)
- [SkillMAS: Skill Co-Evolution with LLM-based Multi-Agent System](https://mlsanity.com/paper/2605.09341/) (2026)
- [Demystifying Agent Skills: Why They Work-Until They Don't](https://mlsanity.com/paper/2608.14036/) (2026)
- [Right Family, Wrong Skill: Evaluating Risk Exposure in Agent Skill Retrieval](https://mlsanity.com/paper/2606.10388/) (2026)
- [The Scaling Laws of Skills in LLM Agent Systems](https://mlsanity.com/paper/2605.16508/) (2026)
- [Tools Are Not Islands: Set-Level Tool Retrieval for LLM Agents via Query-Conditioned Hyperedge Prediction](https://mlsanity.com/paper/2607.25718/) (2026)
- [ReCache: Efficient KV Cache Reuse and Compression for Tool-Augmented LLM Agents](https://mlsanity.com/paper/2608.19662/) (2026)
- [Beyond Task-Only Matching: Personalized Skill Routing with Counterfactual Evaluation](https://mlsanity.com/paper/2608.28241/) (2026)
- [Ace-Skill: Bootstrapping Multimodal Agents with Prioritized and Clustered Evolution](https://mlsanity.com/paper/2605.08887/) (2026)
- [Skill or Skip? Learning Selective Skill Invocation in Agentic Tasks via Dual-Granularity Preference Learning](https://mlsanity.com/paper/2606.00510/) (2026)
- [Unsupervised Skill Discovery for Agentic Data Analysis](https://mlsanity.com/paper/2606.06416/) (2026)
- [SkillCorpus: Consolidating and Evaluating the Open Skill Ecosystem for Real-World LLM Agents](https://mlsanity.com/paper/2607.15557/) (2026)
- [Progressive Agent Skill Generation via Reinforcement Learning](https://mlsanity.com/paper/2608.01678/) (2026)
- [SkillReason: Reasoning-Enhanced Agent Skill Retrieval for Implicit User Requests](https://mlsanity.com/paper/2608.08640/) (2026)
- [Optimal Skill Selection for LLM Agents with Provable Bicriteria Guarantees](https://mlsanity.com/paper/2608.19993/) (2026)

## Abstract (arXiv)

Reusable skills let LLM agents package task-specific procedures, tool affordances, and execution guidance into modular building blocks. As skill ecosystems grow to tens of thousands of entries, exposing every skill at inference time becomes infeasible. This creates a skill-routing problem: given a user task, the system must identify relevant skills before downstream planning or execution. Existing agent stacks often rely on progressive disclosure, exposing only skill names and descriptions while hiding the full implementation body. We examine this design choice on a SkillsBench-derived benchmark with approximately 80K candidate skills, targeting the practically important setting of large skill registries with heavy overlap. Across representative dense and reranking baselines on this setting, hiding the skill body causes a 37-44 percentage point drop in routing accuracy. Stronger controls show that the missing signal is body-resident rather than a simple length artifact: body-distilled descriptions recover part of the gap, but remain 7-21 points below direct all-field routing, while a metadata-only encoder trained with the same data remains 14.0 points below its all-field counterpart. Motivated by this finding, we present Skillrouter, a compact 1.2B body-aware retrieve-and-rerank pipeline. Skillrouter achieves 74.0% Hit@1 on our benchmark -- the strongest average top-1 routing performance among the baselines we evaluate -- while using 13× fewer parameters and running 5.8× faster than the strongest base pipeline. The ranking gains further generalize to a supplementary benchmark independently constructed from three skill sources. In a complementary end-to-end study across four coding agents, routing gains transfer to improved task success, with larger gains for more capable agents.