An Agent-Oriented Pluggable Experience-RAG Skill for Experience-Driven Retrieval Strategy Orchestration
TLDR
Experience-RAG Skill is an agent-oriented system that dynamically selects optimal retrieval strategies for RAG, improving performance across diverse tasks.
Key contributions
- Introduces Experience-RAG Skill, a pluggable agent layer for dynamic retrieval strategy orchestration.
- Analyzes task context and consults an experience memory to select the most appropriate retrieval strategy.
- Achieves 0.8924 nDCG@10 on NQ, HotpotQA, and SciFact, outperforming fixed single-retriever baselines.
- Demonstrates that retrieval strategy selection can be a reusable agent skill, not hard-coded.
Why it matters
This paper addresses a key limitation in RAG by enabling dynamic, task-specific retrieval strategy selection. It shows that treating retrieval orchestration as a reusable agent skill significantly boosts performance across varied tasks, offering a more flexible and efficient way to build adaptive RAG systems.
Original Abstract
Retrieval-augmented generation systems often assume that one fixed retrieval pipeline is sufficient across heterogeneous tasks, yet factoid question answering, multi-hop reasoning, and scientific verification exhibit different retrieval preferences. We present Experience-RAG Skill, an agent-oriented pluggable retrieval orchestration layer positioned between the agent and the retriever pool. The proposed skill analyzes the current scene, consults an experience memory, selects an appropriate retrieval strategy, and returns structured evidence to the agent. Under a fixed candidate pool, Experience-RAG Skill achieves an overall nDCG@10 of 0.8924 on BeIR/nq, BeIR/hotpotqa, and BeIR/scifact, outperforming fixed single-retriever baselines and remaining competitive with Adaptive-RAG-style routing. The results suggest that retrieval strategy selection can be productively encapsulated as a reusable agent skill rather than being hard-coded in the upper workflow.
📬 Weekly AI Paper Digest
Get the top 10 AI/ML arXiv papers from the week — summarized, scored, and delivered to your inbox every Monday.