Model Context Protocol

Earendil has published a new post about its Pi.dev project addressing its earlier decision not to support MCP, the Model Context Protocol for connecting AI assistants to external tools. The piece, titled 'You Said No MCP', appears to revisit that stance and explain the reasoning behind it. Readers on Hacker News are sharing and discussing the company's position on AI tool integration standards.

The modelcontextprotocol organisation's servers repository, written in TypeScript, provides reference implementations of Model Context Protocol servers. The open-source project lets developers connect AI assistants like Claude to data sources and tools such as file systems, databases and third-party services. It is drawing attention among developers working on AI integrations and agentic applications.

Model Context Protocol, or MCP, is being framed as the standard that lets AI agents plug into a company's APIs and tools. But commentary from The New Stack stresses that the protocol only provides access — it does not govern what an agent should or should not see once connected. That means questions of permissions, filtering and data exposure still fall on developers and platform owners rather than on MCP itself.

Earendil's Pi, an AI coding assistant, previously declined to support the Model Context Protocol on its own website but is now integrating it into its core. The shift has prompted discussion of what MCP is and how the assistant's position on the open protocol changed, drawing interest from developers following AI tooling standards.

A developer has published a guide for building a Model Context Protocol (MCP) server that serves documented basketball player data through a typed Sportmicro API integration, written in TypeScript. The tutorial walks through exposing the player insights endpoint so MCP clients can query structured basketball statistics. It is drawing attention from developers interested in connecting sports data to AI tooling.