Duroxide AI Skills

This folder contains AI skills (context/prompts) that can be installed into AI coding assistants to enable effective development with Duroxide.

Available Skills

Skill	Description	For
duroxide-provider-implementation.md	Implementing custom storage providers	Provider developers

Installing Skills

VS Code with GitHub Copilot

VS Code supports multiple instruction methods:

Agent Skills (recommended for specialized workflows):

1. Create .github/skills/duroxide-provider/ in your project
2. Copy skill file as SKILL.md with this header format:

   ---
   name: duroxide-provider
   description: Guidance for implementing Duroxide storage providers
   ---
   

3. Enable with chat.useAgentSkills setting (preview feature)

Custom Instructions (for project-wide guidelines):

1. Copy skill content to .github/copilot-instructions.md for repository-wide rules
2. Or create .github/instructions/duroxide.instructions.md with glob patterns:

   ---
   applyTo: "src/providers/**"
   ---
   

Claude Code

Claude Code uses CLAUDE.md files for project context:

1. Project-level: Copy skill content to ./CLAUDE.md or ./.claude/CLAUDE.md
2. Modular rules: Copy skill file to ./.claude/rules/duroxide-provider.md
3. User-level: Copy to ~/.claude/CLAUDE.md for all projects

Claude Code automatically loads these files. Use /memory command to view loaded files.

Cursor

Cursor supports rules in .cursor/rules/:

1. Project rules: Copy skill file to .cursor/rules/duroxide-provider.mdc
2. With metadata (.mdc format): Add frontmatter:

   ---
   description: "Duroxide provider implementation guidance"
   globs: ["src/providers/**"]
   alwaysApply: false
   ---
   

3. Simple alternative: Copy to AGENTS.md in project root
4. User rules: Add to Cursor Settings → Rules for global application

Other AI Assistants

Skills are standard Markdown files. Copy relevant sections into your AI assistant's system prompt or context window. The Agent Skills open standard is supported by many tools including Roo Code, OpenCode, Goose, and others.

Skill Structure

Each skill file contains:

• Frontmatter — Metadata (title, version, scope)
• Summary — Quick overview of the skill's purpose
• Key Concepts — Core ideas the AI should understand
• Implementation Guidance — Specific patterns and requirements
• References — Links to detailed documentation