Agent Skills for Designers
Agent Skills became an industry standard almost overnight. Here's what designers need to know about this new way to teach AI repeatable workflows, and why it matters for design practice.
Anthropic keeps releasing open infrastructure that becomes an industry standard almost overnight. Last year it was Model Context Protocol. This month it's Agent Skills.
If you're a designer working with AI tools, this matters. Here's why.
The pattern Anthropic's following
In late 2024, Anthropic released Model Context Protocol as an open standard for connecting AI systems to external tools and data. Within months, Google, Microsoft, AWS, and basically every major platform had adopted it. MCP became the de facto way AI agents connect to databases, APIs, and external services.
On December 18, 2025, they did it again. Agent Skills launched as an open standard - a simple format for teaching AI agents how to do specific tasks consistently. And the same adoption pattern is playing out.
Microsoft integrated it into VS Code and GitHub. OpenAI quietly adopted the exact same specification in ChatGPT and their Codex CLI tool - same file structure, same metadata format, same naming conventions. Cursor, Amp, Goose, and OpenCode are all onboard. Enterprise partners like Atlassian, Figma, Canva, Stripe, Notion, and Zapier launched with skills at release.
The two standards work together. MCP provides the hands - connections to external systems. Agent Skills provides the brain - the procedural knowledge for using those tools well. A database MCP server connects Claude to your data; a skill teaches Claude your specific data model and query patterns.
What skills actually are
Think of a skill as a training manual for an AI agent. It's a folder containing a markdown file called SKILL.md that describes how to complete a specific task, plus any scripts, templates, or reference materials the agent needs.
