Why MCP Was Always a Bad Idea

Modern LLMs have become sophisticated enough to directly call APIs and write scripts without needing the Model Context Protocol (MCP), making the protocol obsolete for current use cases. While MCP was useful when released in November 2024 for connecting AI agents to external services, the rapid advancement of language models—particularly their ability to execute code, reason about complex systems, and handle API calls autonomously—has eliminated the need for the intermediate protocol layer. The ecosystem that developed around MCP to solve problems like context bloat is now unnecessary overhead, as models can efficiently manage these tasks independently.