Using Orchard Core Agent Skills

Orchard Core ships a set of agent skills that help AI coding assistants (such as Claude Code, GitHub Copilot, or any agent that supports the skills convention) perform common Orchard Core development tasks consistently and correctly.

A skill is a self-contained folder that bundles step-by-step instructions, code templates, naming conventions, and reference material for a specific task. When an AI agent loads a skill, it follows the project's recommended workflow instead of guessing, which produces output that matches Orchard Core conventions.

Where the skills live

The skills are stored in the repository under:

.agents/skills/

Each skill is a directory containing a SKILL.md entry point and an optional references/ folder with deeper documentation:

.agents/skills/
├── orchardcore-module-creator/
│   ├── SKILL.md
│   └── references/
│       ├── module-structure.md
│       ├── patterns.md
│       └── examples.md
├── orchardcore-theme-creator/
├── orchardcore-asset-manager/
├── orchardcore-admin-edit-views/
└── orchardcore-tester/

Every SKILL.md starts with YAML front matter that declares the skill's name and a description telling the agent when to use it:

---
name: orchardcore-module-creator
description: Creates new OrchardCore modules with proper structure, manifest, startup, and patterns. Use when the user needs to create a new module, add content parts, fields, drivers, handlers, or admin functionality.
---

Available skills

How to use the skills

The exact mechanics depend on your AI assistant, but the general flow is the same.

1. Open the repository in an agent that supports skills

Clone Orchard Core and open it in your AI coding assistant. Agents that follow the skills convention automatically discover the .agents/skills/ folder. The repository root also contains an AGENTS.md file with project-wide guidance (build commands, conventions, and a pointer to the skills) that most agents read on startup.

2. Describe your task in natural language

You don't invoke a skill by name. Instead, describe what you want and the agent matches your request against each skill's description. For example:

• "Create a new module that adds a Rating content part." → triggers orchardcore-module-creator
• "Add a new dark admin theme." → triggers orchardcore-theme-creator
• "Rebuild the SCSS for this module and watch for changes." → triggers orchardcore-asset-manager
• "Fix the layout of this field's edit view." → triggers orchardcore-admin-edit-views
• "Run the app and test that media upload works." → triggers orchardcore-tester

3. Follow the guided workflow

Once a skill is active, the agent walks through its documented steps — creating the required files, applying the correct naming conventions, registering services in Startup.cs, and pulling additional templates from the skill's references/ files when needed.

4. Read the references directly (optional)

The skills are plain Markdown, so they are also useful to humans. Browse a SKILL.md or its references/ files to learn the recommended structure for a module, theme, or admin view even without an AI agent.

Contributing a new skill

Skills are part of the repository and evolve with the codebase. To add or improve one:

1. Create a folder under .agents/skills/<skill-name>/.
2. Add a SKILL.md with name and description front matter. Keep the description action-oriented and explicit about when the skill applies — this is what the agent matches against.
3. Put the high-level workflow in SKILL.md and move long templates or detailed patterns into a references/ subfolder so the entry point stays focused.
4. Follow the conventions in AGENTS.md and the existing skills.
5. Open a pull request. See Contributing code for the general process.

When you add a skill, also list it in the Available Skills table in AGENTS.md so agents and contributors can discover it.