A governance pattern for self-evolving AI skills
Claude Code Skill pattern paper—interesting theory, but unclear if it ships as a usable tool today.

Standards-as-Markdown that auto-load into Claude, Copilot, Cursor — no plugin engineering required.
Government software developers, policy makers, and organizations with technical standards to enforce across teams
GitHub Copilot Policies / Org Rules · Custom instructions / system prompts for AI tools · Spectral linters / governance-as-code frameworks
Skills are Markdown files that inject domain knowledge into AI coding tools. When a developer starts building an API, the tool automatically loads the relevant standard. No plugins to write, no code. Just structured knowledge in Markdown.
I built a marketplace (https://github.com/MinBZK/overheid-claude-plugins) with now 38 skills covering Dutch government standards. Skills are an open standard (https://agentskills.io/) adopted by Claude Code, GitHub Copilot, Cursor, and others.
The interesting part, certainly for a government: you don't have to be a developer to write a skill. A policy officer who knows the standard can structure it in Markdown and suddenly every developer using these tools gets that knowledge for free.
Claude Code Skill pattern paper—interesting theory, but unclear if it ships as a usable tool today.
No merge conflicts by design since every save creates a unique file.
Claude Code skill forces AI to read the diagram before structural work.
Gamified security training for AI agent skills, but it's pre-attack learning, not production defense.
AI exec team for solopreneurs, but it's persona-stacking over real decision infrastructure.
Modular context folders beat monolithic prompts for scaling AI agent instructions.