An open standard for connecting AI clients to tools, prompts, and resources through shared transports—think a common plug shape so one tool server can work across many hosts and frameworks.
MCP separates how a client talks to capabilities from what those capabilities are. Instead of custom N×M integrations per framework and vendor, you expose capabilities through MCP servers that any compatible client can use.
Why sample plan material emphasizes it
The ecosystem content positions MCP next to orchestration frameworks and provider SDKs: it reduces integration overhead when the same databases, repos, or internal APIs must be reachable from multiple agents or editors. For agent-to-agent discovery and delegation, some roadmaps also include the A2A protocol.