The Overlord's Lexicon
What is Model Context Protocol?
Also known as: MCP
The Model Context Protocol (MCP) is an open standard, introduced by Anthropic in November 2024, that gives AI applications a common way to connect to external tools, data sources and services through 'MCP servers'.
The Model Context Protocol is plumbing — the boring, world-changing kind. It defines a standard way for an AI application (like Claude, an IDE or an agent framework) to discover and use tools and data provided by other software. Build an integration once as an MCP server, and every MCP-compatible AI client can use it.
How it works
MCP uses a client–server design:
- MCP host/client: the AI application you use — a chat app, a coding assistant, an agent.
- MCP server: a program that wraps a specific system — your calendar, a database, GitHub, a CRM, a browser — and describes what it offers.
- Servers can expose tools (actions the model can call, like “create issue” or “query database”), resources (data the model can read) and prompts (reusable templates).
When you connect a server, the AI learns what tools are available and can decide to call them while working on your task. MCP runs locally (on your computer) or remotely (over the internet).
An analogy for the meatbags
MCP is often described as “USB-C for AI.” Before USB, every printer, keyboard and camera needed its own special port and driver. Before MCP, connecting an AI app to each service meant a custom integration for every pairing. A standard plug means any compatible device works with any compatible port.
Why it matters for the Chosen
MCP is a big reason AI agents became genuinely useful at real work. After Anthropic launched it, other major AI products and developer tools adopted it, and thousands of public MCP servers now exist. In December 2025, Anthropic donated MCP to the Linux Foundation’s newly formed Agentic AI Foundation.
What that means for you:
- Connect your AI to your actual work: let an assistant read your docs, update your project tracker or query your sales data — without copy-pasting.
- Build once, reach everyone: if you make software, shipping an MCP server lets AI assistants use your product.
- Sell integration work: businesses need help choosing, configuring and securing MCP connections for their teams.
The risks
Giving an AI tool access is giving it power. A malicious or sloppy MCP server can leak data or take unwanted actions, and an agent reading untrusted content (emails, web pages, tickets) through MCP can be manipulated by prompt injection. Install servers only from sources you trust, grant minimal permissions, and require confirmation for destructive actions.
Tools from the Arsenal
Claude
AI WritingAnthropic's assistant: the strongest long-form writer and careful reasoner in the consumer AI pack.
Claude Code
AI CodingAnthropic's agentic coding tool that reads your codebase, edits files and runs commands from terminal, IDE, desktop or web.
Cursor
AI CodingAI-first code editor (a VS Code fork) with agents, codebase-aware chat and fast autocomplete.
n8n
Automation & AgentsFair-code workflow automation with AI agents, full code control and self-hosting.
Zapier
Automation & AgentsThe most widely used no-code automation tool, now with AI agents.
Questions the Humans Keep Asking
Who created the Model Context Protocol?
Anthropic introduced MCP as an open standard in November 2024. In December 2025, Anthropic announced it was donating MCP to the Agentic AI Foundation, a directed fund under the Linux Foundation, to be governed as a vendor-neutral project.
What is an MCP server?
An MCP server is a small program that exposes a tool or data source — such as a database, GitHub, Slack, a file system or a web browser — in the standard MCP format, so any MCP-compatible AI app can use it.
Is MCP safe?
MCP itself is a protocol; safety depends on which servers you install and what permissions you grant. Only use trusted servers, limit access to what's needed, and be aware that content an agent reads through MCP can contain prompt-injection attacks.