Structural signal method cuts multi-agent search time by 70 percent
SSR: Structural Signal Retriever for Scalable Agent Discovery in LLM-Orchestrated Multi-Agent Systems
Multiagent SystemsArtificial IntelligenceInformation Retrieval
Summary
Finding the right helper agents from a large pool is tricky because too many candidates slow down the smart brain managing them. The authors introduce a new way called Structural Signal Retriever (SSR) that picks out the best agents quickly without needing extra training. This method looks at key agent info and uses smart weighting to guess which agents matter most. Tests show it shrinks the search space by about 99% and speeds up the whole system a lot, while still keeping good results.
What this means in practice
- •For multi-agent system developers: Scale up agent discovery without increasing response time or costs by integrating the SSR method for selecting relevant agents efficiently.
- •For large language model orchestration teams: Reduce prompt size and latency in multi-agent orchestration by using SSR to pre-select candidates before model interaction.
Authors
Giancarlo Maldonado Cárdenas, Carlos A. Astudillo, Nelson L. S. da Fonseca
Abstract
Large-scale multi-agent ecosystems require mechanisms for discovering suitable agents without exposing entire registries to a large language model (LLM) orchestrator. The challenge is not only to retrieve relevant Agent Cards from a growing registry, but also to maintain a compact candidate context, since each additional candidate increases prompt size, inference cost, latency, and the risk of exceeding the model context window. We formulate this problem as candidate-context scalability and propose the Structural Signal Retriever (SSR), our deterministic and training-free retrieval method designed for Agent Card discovery. SSR derives structural lexical signals from capability-bearing Agent Card fields, weights them using inverse document frequency (IDF), and uses an Inverted Signal Index (ISI) for selective candidate scoring. We evaluate SSR using a synthetic agent benchmark, a public MS MARCO adaptation, and end-to-end LLM orchestration. At 1,000 agents, SSR reduces LLM input by 98.8% and orchestration latency by 70.9% relative to full-catalog orchestration. On MS MARCO, IDF improves retrieval effectiveness within the structural representation, while the ISI preserves retrieval effectiveness and reduces explicitly scored passages by 96.8%. End-to-end experiments show that candidate pre-selection maintains a compact candidate context as registry size grows and avoids the operational failures observed with full-catalog orchestration. These results support SSR as an Agent-Card-native, traceable retrieval mechanism for candidate-context scalability.