paperarXivTrust 82 · PrimaryPublished 4d agoLive · 3d ago
Linguistic Firewall: Geometry as Defense in Multi-Agent Systems Routing
The rapid integration of Large Language Models (LLMs) has driven the evolution of Multi-Agent Systems (MAS), where specialized agents collaborate to execute complex workflows. Effective orchestration in these environments requires robust routing mechanisms to efficiently allocate tasks to the most suitable agent. However, existing routers fundamentally rely on unverified proxies, ranging from textual self-descriptions to static surrogate representations, to gauge an agent's competence. This reliance on non-empirical data creates a critical gap between an agent's projected profile and its actua
Lineage graph
Paper → model → repo connections mined from source citations (Tier-1 exact match).
Implements
Covers
Has model
Implements (incoming)
Covers (incoming)
Related across the graph
repoBoundaryML/bamlrepodeepset-ai/haystackreponaberbabammm34343/llm-task-orchestratornewsShow HN: Smart model routing directly in Claude, Codex and CursornewsPrompt injection is exploiting enterprise AI's biggest design flaws by targeting agents, RAG pipelines and model routersrepodoublewordai/control-layerrepomarciok/gustrepovalorisa/Advanced-Multi-Agent-Orchestratorrepoenvoyproxy/ai-gatewayrepoImFeH2/flowentmodelAgentCore-8BnewsIEEE Rolls Out Large Language Models Virtual Training CoursenewsA system-level approach to prompt injection: separating instruction and data channels in LLM agents [P]repollm-d/llm-d-routerrepolangchain-ai/langgraphjsrepoagent-tools