MCP (Model Context Protocol)
An open standard developed by Anthropic in 2024 that defines how AI models connect to external tools, databases, and services. MCP gives AI agents a consistent interface for reading files, querying systems, and triggering actions — without custom integration code for every tool.
What is MCP (Model Context Protocol)?
MCP — Model Context Protocol — is an open standard released by Anthropic in late 2024. It defines a universal interface for connecting AI models to the tools and data sources they need to act: databases, file systems, APIs, ERP tables, calendar systems, CRMs. Before MCP, every AI integration required bespoke glue code. MCP standardises the handshake.
Think of it like USB for AI. Instead of building a custom connector every time you want an AI agent to read a purchase order from your ERP or query a supplier database, you write one MCP server — and any compatible AI model can use it immediately.
How MCP Works
MCP separates the AI model (the reasoning engine) from the tools it can use. An MCP server exposes a set of tools, resources, and prompts to any connected client. The AI model calls these tools during a session — fetching live data, writing records, triggering downstream processes — without the model itself needing to know anything about the underlying system's architecture.
Tools: Actions the AI can invoke (e.g., query stock level, create a PO line)
Resources: Data the AI can read (e.g., a supplier master file, an open order list)
Prompts: Pre-built instruction templates for common tasks
MCP in Operations
For operations and IT teams evaluating AI infrastructure, MCP changes the build vs. integrate calculus. Instead of maintaining a patchwork of one-off AI connectors for your ERP, WMS, and supplier portal, you build MCP servers once per system and expose them to all your AI agents. A procurement agent that checks stock, a finance agent that validates invoices, and a logistics agent that tracks shipments can all draw on the same MCP-connected data layer.
MCP is already supported natively by Claude (Anthropic), and adoption across major AI platforms accelerated through 2025. If your team is planning an AI rollout, MCP compatibility should be on your vendor evaluation checklist.