Today, we’ve released to the community two powerful new tools that make it easy to use modern AI with NetBox: a Model Context Protocol server for NetBox to give LLMs the ability to directly interact with NetBox data, and multiple variations of llms.txt for educating LLMs with context about NetBox. Read on below to learn more about what these tools are, how they supercharge LLMs for use with NetBox, and how you can get started with them right now to amp up your NetBox workflows with AI.
Teaching LLMs about NetBox
One of the distinct advantages of working with NetBox is that modern AI language models already know NetBox well because of NetBox’s deep documentation, massive community and install base, and industry standard data models for network and infrastructure documentation. No other network and infrastructure management tool is as “well trained” into modern LLMs as NetBox.
Despite that, modern LLMs have knowledge cutoffs and NetBox is always evolving. Plus, LLMs can’t directly interact with systems like NetBox to unlock AI-driven interaction with your NetBox data. These are the two problems we’re solving with today’s releases: delivering tools for teaching LLMs current, complete context about the latest NetBox versions, and also tools for LLMs to directly interact with your data in NetBox for knowledge exploration and analysis.
The NetBox MCP server: connect AI directly into NetBox
LLMs are great partners for working with NetBox. But what about when you want to connect your own real NetBox data with AI? Anthropic released the open source Model Context Protocol (MCP) last year for just this reason. MCP is a standard way for any application to connect to an LLM. With MCP, LLMs can now find and use third party APIs and services.
Today we’re releasing the open source NetBox MCP server, which enables read-only interactions between AI tools that support MCP and any NetBox instance, even private NetBox instances that aren’t accessible online.
What does this mean, practically? Let’s look at a quick example. The video below shows an interaction with the NetBox demo dataset, showing a simple AI-led data exploration. We ask the AI for help with one task: telling us what’s in our Syracuse datacenter. With the NetBox MCP server and the LLM’s inherent knowledge of the NetBox data model, the AI explores the data in NetBox and delivers us a detailed summary.
The NetBox MCP server connects AI directly with your NetBox data
With the NetBox MCP server, AI becomes a powerful partner for novel exploration and reporting of your network and infrastructure data. AI can assist in planning and optimization tasks, accelerate troubleshooting efforts, and much more.
To get started, check out the NetBox MCP server on GitHub. Get the code and read the docs for how to set up the NetBox MCP server and connect it with an LLM.
LLMs.txt: NetBox documentation for LLMs
Modern language models are incredibly powerful given the right context. They know NetBox well, but it’s helpful to fully instruct them on “current state” documentation of the NetBox ecosystem. Recently, a simple emerging standard for delivering comprehensive context to LLMs has appeared: llms.txt.
An llms.txt file encodes content in Markdown suitable for consumption by AI so you can quickly and easily give language models all the context they need about a particular topic – in our case, NetBox and how to work with it. With llms.txt, you can easily show language models what NetBox is and teach them all of NetBox’s latest capabilities.
We’ve published not just one, but three llms.txt files for the NetBox community:
- NetBox llms.txt – at netboxlabs.com/docs/llms.txt – A lightweight index of all the key documentation for NetBox to help LLMs know where to look
- NetBox llms-full.txt – at netboxlabs.com/docs/llms-full.txt – A complete set of NetBox documentation content for LLMs to come fully up to speed on NetBox in a single shot (be aware: this will use a lot of tokens)
- NetBox REST API llms-full.txt – at netboxlabs.com/docs/llms-api-full.txt – A separate complete set of NetBox REST API docs, suitable if you’re working with LLMs to integrate with the NetBox REST API
The files at these URLs will be updated as NetBox evolves, and will grow to include other important content to teach AI about NetBox as well. You can get started with the NetBox llms.txt files super easily. If you’re using ChatGPT or another chat-based AI tool, upload the file to a project as context. If you’re using Cursor, index the files in your project. And with that, you’ve immediately created an AI “NetBox expert” to help you in your use of NetBox.
AI in network and infrastructure management is just beginning
AI is already having a big impact on the way networks and infrastructure are managed. Just this quarter, we’ve delivered multiple NetBox AI features for NetBox Enterprise and NetBox Cloud customers, like intelligent AI enrichment for alerts and tickets, and AI-driven NetBox upgrade risk analysis. We’ve got many more NetBox AI features on the way, and today’s community releases only amplify the benefits of using NetBox as AI becomes more prevalent in network and infrastructure management this year and beyond.
We want to hear from you. Do you have questions or feedback about llms.txt or the NetBox MCP server? Join us on NetDev Chat in #netbox and share.
And if you’re wondering how to get started with AI for network and infrastructure management, or want to get a sense for what this all means and what your team needs to be doing now to prepare for an AI-driven world, join us next week for our live webinar: Network and Infrastructure AI is Here: How Teams Should Prepare. We’ll share more details of our vision for network and infrastructure AI, share more details and examples of AI features in NetBox today, and share what we’re learning as we engage with teams around the world on AI use cases.
