AT Protocol Model Context Protocol server for social web interactions.
Providing version-specfic context to an LLM using an MCP can greatly improve productivity and reduce cost.
Workday's CEO called AI agent startups "parasites." Linear shipped a native AI agent that understands software development workflows. These are two ends of a spectrum, with most of the industry stuck in the middle — exposing data through MCP without embedding any intelligence. The companies that win will be the ones that make AI native to their domain workflow.
LLMs have statistical momentum: even when they know a new standard like Streamable HTTP exists, they often revert to the legacy patterns they saw most in training. Here is how to use "strong anchors" and "zero-prompt pruning" to keep your agentic systems from being haunted by 2024.
MCP Apps adds an official extension for rendering interactive UIs inside MCP clients, building on MCP-UI and OpenAI's Apps work to move beyond plain text.
After three years at Microsoft working on security and MCP, I'm joining Anthropic as a Member of Technical Staff to help grow the MCP ecosystem.
Wir sind aktuell dabei weiter an einem MCP Server und dazugehörige Erweiterung in Magento 2 zu arbeiten. Der MCP Server ermöglicht es, dass LLMs per MCP gewisse Endpoints zur Verfügung stellen kann. Diese Endpoints erlauben den Zugriff auf Daten im Shop (später evtl. auch Manipulation der Daten). Der MCP-Server wiederum baut eine Verbindung zu Magento...
Year-in-review piece on a 2025 spent largely on Model Context Protocol work, covering spec milestones, ecosystem shifts, and what's likely next.
A core maintainer's reflection on a year of Model Context Protocol, coinciding with the 2025-11-25 specification release and a look at what's next.
How Client ID Metadata Documents replace Dynamic Client Registration in the MCP authorization spec, and how Visual Studio Code will surface CIMD to users.
Recap of the MCP maintainers gathering in New York - how the contributor community coordinates async work on auth, security, registries, and docs.
VS Code Insiders adds a consent prompt before installing a third-party MCP server, blocking silent one-click writes to the editor's MCP configuration.
VS Code now supports MCP elicitations from the 2025-06-18 spec, letting servers prompt users with structured native dialogs instead of guessing at parameters.
Why most teams should let an API gateway handle the new MCP Protected Resource Metadata flow instead of hand-rolling OAuth in their own server code.
Notes from a month of MCP work spanning Microsoft Build, the MCP Developer Summit, AI Engineer World's Fair, and ongoing security and spec contributions.
Why AI agents act as a force multiplier for engineers, why the job-replacement framing misses the point, and how to think about agent-based workloads.
Mitigating confused deputy attacks in remote MCP servers by enforcing per-client consent prompts and proxying Dynamic Client Registration through Azure APIM.
This isn't an exhaustive or fully baked idea yet, but I've been noticing a trend with MCP servers -- they love to just yeet a bunch of JSON at an LLM. I think this is well-intentioned but not super optimal.
Parent WAM authentication dialogs to MCP clients like VS Code and Claude Desktop by traversing the process hierarchy to find a usable window handle.
Building an Entra ID-protected remote MCP server with Azure API Management as the auth gateway and Azure Functions handling the protocol implementation.
MCP brings back classic security problems around local execution and remote authentication, and the industry seems determined to ignore lessons already learned.
Walkthrough of the MCP authorization RFC that adds Protected Resource Metadata so servers can delegate auth to GitHub, Entra ID, and other existing IdPs.
OpenSpartan Forerunner is a Model Context Protocol server that exposes Halo Infinite stats and gear data to MCP clients like VS Code and Claude Desktop.