Is MCP Good Yet?

The article evaluates the implementation status of the Model Context Protocol (MCP) across various AI development platforms and clients, comparing features like protocol versioning, elicitation (user input during tool calls), tasks support, and code mode functionality. Most advanced features remain partially implemented or absent across Cursor, Claude Code, Grok, and other platforms, with the 2026-07-28 protocol revision unevenly adopted and key capabilities like long-running tasks still lacking standardized support. The assessment reveals significant gaps between the MCP specification and actual shipping implementations across major AI coding tools.