Hi Zabbix devs,

AI tooling is accelerating globally, and many systems are building in native support for open AI interfaces. I’d like to propose Zabbix explores supporting the Model Context Protocol (MCP), a standard that allows AI agents to interface with systems using a structured, declarative protocol.

With MCP support, AI models (like GPT-4, Claude, etc.) could interact with Zabbix in a standardized way—asking questions, reading triggers, exploring item history, or suggesting remediations.

Some example use cases:

• AI assistants that explains what caused a trigger to fire.

• LLM-based automation tools that read Zabbix states and performs actions via an orchestrator.

• Teams using LangChain or Flowise could easily pull Zabbix data into workflows.

Supporting MCP early on would position Zabbix as an AI-integrated platform and avoid vendor lock-in by relying on open standards. Seems to fit the Opensource standards of Zabbix well.