The FOCUS MCP Server gives AI assistants direct, read-only access to the FOCUS specification as a live knowledge source. Practitioners can ask questions in natural language and get answers direct from the specification and content without continually updating supporting context.

FOCUS MCP resources available:

• Specification: specification details, filterable by version
• Column Library: retrieve FOCUS column definitions, including its category, dataset, and version history
• Use Case Library: 100+ use cases with related capabilities, KPIs, queries, columns, and personas
• SQL Queries: baseline queries tied to use cases, ready to adapt to your environment
• Provider Directory: which data generators support FOCUS and how to get started with each
• Supporting Information: FAQs, adoption guidance, and other reference content from focus.finops.org

How It Works

The server implements the Model Context Protocol (MCP), the open standard for connecting AI assistants to external data sources. Any MCP-compatible client (Claude, Cursor, Windsurf, VS Code with Copilot, and others) can connect to it. The assistant discovers available tools at connection time and calls them as needed during a conversation.

Practitioners OR Agents can get the accurate answers to questions like:

• “Which FOCUS columns do I need to reconcile invoices to usage?”
• “Show me the use case for analyzing commitment coverage, and the related SQL.”
• “What changed between v1.2 and v1.3 for the BilledCost column?”
• “Explain the difference between ServiceProviderName and HostProviderName.”

Why This Matters

The FOCUS spec is large and growing. Navigating 77+ columns, 104+ use cases, and multiple spec versions takes time, especially when you’re trying to write a query or build a dashboard. The MCP server puts that reference material inside the tools practitioners already use, so the spec becomes something you integrate in rather than something you look up.