Get Started
Home
Topics
Search
Library
Research questionHow can LLM agents retrieve relevant skills from large, noisy libraries under context and latency constraints?As skill libraries grow, user requests may contain only small signals about which reusable skill is appropriate, while queries and skill descriptions can be long and noisy. Retrieval must identify suitable skills without exceeding the agent’s context or latency budget.
AI
AI Agents
Evaluation & Benchmarks
Information Retrieval
Natural Language Processing
Latest papersRecent research connected to this question, newest first.SkillRet: A Large-Scale Benchmark for Skill Retrieval in LLM AgentsThe source studies retrieval of reusable agent skills from 16,129 public skills organized with semantic tags in a two-level taxonomy. Its evidence includes 63,259 training samples and 4,392 evaluation queries with disjoint skill pools, comparing diverse retrievers and analyzing behavior on long, noisy queries; the reported results are benchmark-based rather than evidence of deployment performance beyond those settings.research paper · Sep 1, 2026
Related questions
How can LLM agents reuse procedural skills across diverse long-horizon tasks without generic or instance-bound memory?How can LLM agents jointly adapt reasoning policies and hierarchical skill libraries during reinforcement learning?How can LLM agents reuse execution traces without losing temporal and outcome-dependent behavior?How can agent runtimes avoid context poisoning and latency from growing histories during long-horizon skill execution?