Orb strives to be more than just the world’s best Internet and network experience monitoring solution. We don't want to lock you into a specific set of tools, use cases, and workflows. Instead, we have built a platform with run-anywhere agents, flexible SDKs, full control of your data, and access to the same powerful APIs that back Orb Cloud to enable integrating any aspect of the Orb experience into your own products and services. Given this flexibility, Orb customers have already integrated Orb data and APIs into their agentic AI workflows. Today, we are making this process frictionless with the release of our MCP for Orb Cloud.

What is MCP?

Model Context Protocol, or MCP for short, is an open protocol that lets AI assistants connect to external tools, data sources, and services in a standardized way. Every organization has their own unique requirements and policies around the use of AI: which vendors are used, which models, where inference is run, etc. MCP allows Orb to support integration everywhere from major frontier AI labs to popular open source tooling.

Orb’s MCP server interfaces with Orb Cloud to allow you to check the current status of your Orbs, analyze and visualize historical data, get recommendations for improving your internet experience, trigger tests, and more. To see a list of available capabilities, visit the documentation.

Use cases

We’ve been using the MCP for Orb Cloud over the past few months to streamline support for Orb customers that have worked with us during our beta period. When customers want to zero-in on specific events or specific times across multiple time zones on a large number of deployed devices, generating reports from traditional dashboards and analytics tools can be tedious and time-consuming. Claude Cowork combined with our MCP server can make short work of a comprehensive report with key events visualized and an executive summary.

The tool is also a great option for customer care and IT help desk teams, by equipping them with Orb’s wealth of data and insights, the overall context of the model, and your own company-specific augmentations. The combination of experience over time data, Wi-Fi dimensions, and the ability for the MCP to remotely control Orbs directly, means that agents can help your team create better outcomes for your users and subscribers in less time.

While the MCP today is not designed for direct integration with your customer-facing agents for self-care scenarios, it is immensely useful for prototyping and experiencing the value of Orb data for this use case ahead of a production roll-out using Orb APIs or Datasets.

Can I use it without a paid plan?

Yes! All Orb users have access to an included 500 tool calls per month. Free plan users can access near-real-time insights from their Orbs as well as the last 72 hours of Scores Datasets. Historical Responsiveness, Web Responsiveness, Speed, and Wi-Fi Datasets, access to 90 days of history, and higher tool call quotas are reserved for Orb Pro plans.

Getting started

The MCP for Orb Cloud is now generally available. AI tools and platforms that support remote MCP servers via API authentication or OAuth are generally supported. More details and instructions for setting up OpenAI Codex, Claude Desktop, Claude Code, Copilot Studio, and OpenCode are available in the Orb docs. Enable the Orb MCP server in your favorite AI platform today!