Your PC, Your Rules: AI Agents Meet Windows Personalization | OD858

Samantha Song demonstrates an agent-driven approach to Windows personalization where a user can describe a desired “vibe” in natural language (for example, a late-night coding theme) and have the system apply concrete Windows personalization changes.

Overview

The session explains how natural-language prompts can be mapped to Windows personalization APIs using MCP servers, with a clear flow from:

It also calls out practical considerations such as scope, permissions, and rollback.

What the session covers

Windows personalization scenarios driven by prompts

The video describes using MCP servers to drive Windows personalization features such as:

Architecture: from intent to API execution

Key ideas highlighted in the description and chapter list include:

Dynamic themes and lighting deep dive

The session includes a deeper look at Dynamic Lighting and RGB effects, including:

Building the skill with GitHub Copilot

A segment focuses on implementing the personalization skill using GitHub Copilot, including:

Session context and resources

Chapters (from the video description)