← Back to Glossary
Glossary

Model Context Protocol (MCP)

Model Context Protocol (MCP) is an open standard introduced by Anthropic in 2024 that defines how AI models connect to external tools and data sources through a unified interface. Before MCP, every integration between a model and an external system—a database, a file system, an API—required custom glue code. MCP standardizes this with a client-server architecture: the AI application is the client, and each data source or tool runs an MCP server that exposes its capabilities in a format any MCP-compatible model can understand. Think of it as USB for AI integrations. The protocol handles tool discovery, authentication, and structured input/output so developers build the connector once and any MCP client can use it. MCP matters because the value of AI systems increasingly depends on what they can access and act on, not just what they know from training.

Related terms:

Generative Engine Optimization

Generative engine optimization (GEO) is the practice of structuring content so AI systems—such as ChatGPT, Perplexity, Google AI Overviews, and Bing Copilot—cite and surface it when answering queries. Unlike traditional SEO, GEO prioritizes clear, BLUF-style definitions, structured data markup, and authoritative sourcing to ensure LLM retrieval and citation while avoiding filler.

AI Evaluation

AI evaluation is the practice of systematically measuring an AI system’s performance against defined criteria—accuracy, latency, cost, safety, and user satisfaction—using representative test datasets, business-outcome metrics, and automated pipelines before and after deployment. Without it, organizations risk flying blind, mistaking demo success for reliable production performance.

System Prompt

A system prompt is an invisible set of instructions given to a language model—defining its persona, constraints, output format, and behavioral rules—and occupies the “system” role in APIs like OpenAI and Anthropic. It shapes every response by encoding business logic and is the most efficient way to control model behavior.