MCP-Diag: A Deterministic, Protocol-Driven Architecture for AI-Native Network Diagnostics

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Lodha, Devansh, Panchal, Mohit, Kulkarni, Sameer G.
Formato: Preprint
Publicado: 2026
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866912862179426304
author Lodha, Devansh
Panchal, Mohit
Kulkarni, Sameer G.
author_facet Lodha, Devansh
Panchal, Mohit
Kulkarni, Sameer G.
contents The integration of Large Language Models (LLMs) into network operations (AIOps) is hindered by two fundamental challenges: the stochastic grounding problem, where LLMs struggle to reliably parse unstructured, vendor-specific CLI output, and the security gap of granting autonomous agents shell access. This paper introduces MCP-Diag, a hybrid neuro-symbolic architecture built upon the Model Context Protocol (MCP). We propose a deterministic translation layer that converts raw stdout from canonical utilities (dig, ping, traceroute) into rigorous JSON schemas before AI ingestion. We further introduce a mandatory "Elicitation Loop" that enforces Human-in-the-Loop (HITL) authorization at the protocol level. Our preliminary evaluation demonstrates that MCP-Diag achieving 100% entity extraction accuracy with less than 0.9% execution latency overhead and 3.7x increase in context token usage.
format Preprint
id arxiv_https___arxiv_org_abs_2601_22633
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle MCP-Diag: A Deterministic, Protocol-Driven Architecture for AI-Native Network Diagnostics
Lodha, Devansh
Panchal, Mohit
Kulkarni, Sameer G.
Networking and Internet Architecture
Artificial Intelligence
68M10
C.2.3; I.2.7
The integration of Large Language Models (LLMs) into network operations (AIOps) is hindered by two fundamental challenges: the stochastic grounding problem, where LLMs struggle to reliably parse unstructured, vendor-specific CLI output, and the security gap of granting autonomous agents shell access. This paper introduces MCP-Diag, a hybrid neuro-symbolic architecture built upon the Model Context Protocol (MCP). We propose a deterministic translation layer that converts raw stdout from canonical utilities (dig, ping, traceroute) into rigorous JSON schemas before AI ingestion. We further introduce a mandatory "Elicitation Loop" that enforces Human-in-the-Loop (HITL) authorization at the protocol level. Our preliminary evaluation demonstrates that MCP-Diag achieving 100% entity extraction accuracy with less than 0.9% execution latency overhead and 3.7x increase in context token usage.
title MCP-Diag: A Deterministic, Protocol-Driven Architecture for AI-Native Network Diagnostics
topic Networking and Internet Architecture
Artificial Intelligence
68M10
C.2.3; I.2.7
url https://arxiv.org/abs/2601.22633