Model Context Protocol (MCP)
Also known as: Model Context Protocol
An open protocol (from Anthropic) for letting LLMs call tools and read resources from any compatible server.
An open protocol for connecting LLMs to external tools and data sources. Standardizes tool definitions, invocation, and response formats so agents can interact with arbitrary APIs.
In practice
Becoming the default plug for agent tooling. Senior candidates should understand server/client roles and how it compares with OpenAI function calling.
How it compares
MCP is the wire protocol; an agent is the LLM-driven system that USES MCP (or function calls) to act.
Comparisons that include Model Context Protocol (MCP)
Related topics
Questions that mention this term
Related terms
Prompt Engineering
Crafting the input text so the model produces the behavior you want.
AI Agents
LLMs that loop: plan → call tools → observe results → repeat until done.
Chain-of-Thought (CoT)
Ask the model to think step by step before answering. It boosts accuracy on reasoning tasks.
In-Context Learning (ICL)
Show the model a few examples in the prompt and it learns the pattern, no fine-tuning needed.
Function Calling
The model emits structured JSON specifying a tool call instead of free text; the host executes it.
Prompt Injection
Attacker text in user input or retrieved content overrides the model's system prompt or hijacks tool calls.