NetBox Labs added infrastructure-design templates and AI agents at its Florida conference. Agents begin with read-only access; proposed changes move through a separate branch, architect review and rule validation. The announcement extends the company’s infrastructure platform with reusable plans, dashboards and tools that connect customers’ own agents to its records of networks and data centers.
NetBox Labs, the commercial steward of the open-source network inventory tool NetBox, announced new platform capabilities at its first NetBox Evolve conference in Florida on October 13. AI agents are among the additions.[1], [2]
Chief executive Kris Beevers described agents that start with read-only access, with proposed infrastructure changes routed through architect review rather than immediately applied to operational equipment.[1]
A proposed change begins in NetBox’s model of the intended infrastructure and runs on a separate branch. NetBox Validation checks it against operating and compliance rules and sends feedback to the agent. The pre-built agents cover change-request review, equipment lifecycle and version compliance, data quality and data import. A managed server using Model Context Protocol, the mechanism connecting models with external tools, gives customers’ own agents access to NetBox infrastructure information.[1]
NetBox Designs is in public preview. Teams can define a site, rack or internet-address plan, specify its parameters and reuse it down to individual components and configuration. Beevers named repeated AI data center racks and standard retail-site technology as examples. NetBox Analytics supplies dashboards for infrastructure questions. NetBox Asset Lifecycle follows planned objects through materials lists, purchasing, shipping and installation. These additions extend the Infrastructure Intelligence Platform introduced in June.[1]
