# Infrahub AI Skills AI skills for developing with [Infrahub](https://github.com/opsmill/infrahub), the infrastructure data management platform by OpsMill. These skills provide structured guidance — covering schema design, data population, validation checks, generators, transforms, and menu customization — to any AI coding assistant that supports custom instructions or context files. The skills are written in plain Markdown with lightweight YAML frontmatter and are easy to adapt to whichever AI tool you use. ## Installation ### Copy Skills into Your Repository The simplest and most universal approach — works with any AI tool. Copy the \`skills/\` directory into your project so the AI assistant can discover the files directly: \`\`\`bash # Clone the skills repository git clone https://github.com/opsmill/infrahub-skills.git # Copy all skills into your Infrahub project cp -r infrahub-skills/skills /path/to/your-infrahub-project/ # Clean up rm -rf infrahub-skills \`\`\` Or copy only the skills you need: \`\`\`bash # Example: copy just schema and object creator skills mkdir -p /path/to/your-infrahub-project/skills cp -r infrahub-skills/skills/schema-creator /path/to/your-infrahub-project/skills/ cp -r infrahub-skills/skills/object-creator /path/to/your-infrahub-project/skills/ cp -r infrahub-skills/skills/common /path/to/your-infrahub-project/skills/ \`\`\` > **Note:** Always include \`skills/common/\` when copying individual skills — it contains shared references (GraphQL queries, \`.infrahub.yml\` format, git integration rules) that all skills depend on. ### Claude Code (Plugin) If you prefer the plugin approach, the skills are automatically available across all your Infrahub projects without copying files into each one. #### Option 1: Via OpsMill Marketplace (Recommended) Run these commands inside Claude Code: \`\`\`bash # Add the OpsMill marketplace /plugin marketplace add opsmill/claude-marketplace # Install the Infrahub plugin from the marketplace /plugin install infrahub@opsmill \`\`\` Or add to your \`~/.claude/settings.json\`: \`\`\`json \{ "extraKnownMarketplaces": \{ "opsmill": \{ "source": \{ "source": "github", "repo": "opsmill/claude-marketplace" \} \} \}, "enabledPlugins": \{ "infrahub@opsmill": true \} \} \`\`\` #### Option 2: Via Git Clone and install directly from the repository: \`\`\`bash # Clone the repository git clone https://github.com/opsmill/infrahub-skills.git # Install the plugin (run inside Claude Code) /plugin install ./infrahub-skills \`\`\` #### Option 3: Local Path If you have the plugin locally, run inside Claude Code: \`\`\`bash /plugin install /path/to/infrahub-skills \`\`\` ## Skills ### Schema Creator **Name:** \`infrahub-schema-creator\` Create, validate, and modify Infrahub schemas for infrastructure data management. **Capabilities:** - Create new schemas from natural language requirements - Design nodes with attributes, relationships, and generics - Set up hierarchical location trees and component/parent patterns - Configure display properties (human_friendly_id, display_label) - Validate schemas and plan migrations **Documentation:** \`skills/schema-creator/\` -- SKILL.md, reference.md, examples.md, validation.md, rules/ --- ### Object Creator **Name:** \`infrahub-object-creator\` Create and manage Infrahub object data files for populating infrastructure data. **Capabilities:** - Create YAML data files for devices, locations, organizations, modules - Set up hierarchical data (location trees, tenant groups) - Reference related objects across files - Manage component children (interfaces, modules, bays) - Organize files for correct dependency load order **Documentation:** \`skills/object-creator/\` -- SKILL.md, reference.md, examples.md, rules/ --- ### Check Creator **Name:** \`infrahub-check-creator\` Create validation checks that run in proposed change pipelines. **Capabilities:** - Write Python validation logic (InfrahubCheck class) - Create GraphQL queries for data fetching - Build global and targeted checks - Register checks in .infrahub.yml - Debug check failures **Documentation:** \`skills/check-creator/\` -- SKILL.md, examples.md, rules/ --- ### Generator Creator **Name:** \`infrahub-generator-creator\` Create design-driven generators that automatically create infrastructure objects. **Capabilities:** - Build generators that create objects from design definitions - Implement idempotent create-or-update workflows (allow_upsert) - Set up target groups and GraphQL queries - Configure automatic cleanup of stale objects - Register generators in .infrahub.yml **Documentation:** \`skills/generator-creator/\` -- SKILL.md, examples.md, rules/ --- ### Transform Creator **Name:** \`infrahub-transform-creator\` Create data transforms that convert Infrahub data into different formats. **Capabilities:** - Build Python transforms (InfrahubTransform class) - Create Jinja2 template-based transforms - Generate device configs, CSV reports, inventory exports - Connect transforms to artifacts for automated output - Hybrid Python + Jinja2 patterns **Documentation:** \`skills/transform-creator/\` -- SKILL.md, examples.md, rules/ --- ### Menu Creator **Name:** \`infrahub-menu-creator\` Create custom navigation menus for the Infrahub web interface. **Capabilities:** - Design navigation menus with nested hierarchies - Organize node types into logical groups - Configure icons (MDI icon set) and labels - Link menu items to schema node list views **Documentation:** \`skills/menu-creator/\` -- SKILL.md, rules/ --- ### Common References Shared documentation and rules referenced by all skills. **Contents:** - \`graphql-queries.md\` -- GraphQL query writing reference - \`infrahub-yml-reference.md\` -- .infrahub.yml project configuration - \`rules/\` -- Shared rules (git integration, caching gotchas) **Location:** \`skills/common/\` ## Using with Other AI Tools These skills follow the [Agent Skills](https://platform.claude.com/docs/en/agents-and-tools/agent-skills/overview) format — each skill is a directory with a \`SKILL.md\` entry point (with \`name\` and \`description\` frontmatter) plus supporting reference files. Any AI coding tool that supports skills can use them directly. For tools that don't support skills natively, copy the \`skills/\` directory into your project and point the tool at the relevant files. ### GitHub Copilot Copy the \`skills/\` directory into your repo, then reference the skills from a Copilot instructions file: \`\`\`bash cp -r skills/ /path/to/your-infrahub-project/skills/ \`\`\` Create \`.github/instructions/infrahub.instructions.md\` to point Copilot at the skills: \`\`\`markdown --- applyTo: '**/*.py,**/*.yml,**/*.yaml' --- For Infrahub development guidance, refer to the skill files in skills/. \`\`\` See [GitHub Copilot custom instructions docs](https://docs.github.com/en/copilot/how-tos/configure-custom-instructions). ### Cursor Copy the \`skills/\` directory into your repo, then create a Cursor rule to reference them: \`\`\`bash cp -r skills/ /path/to/your-infrahub-project/skills/ \`\`\` Create \`.cursor/rules/infrahub.mdc\`: \`\`\`markdown --- description: Infrahub development guidance globs: ["**/*.py", "**/*.yml", "**/*.yaml"] alwaysApply: false --- For Infrahub development guidance, refer to the skill files in skills/. \`\`\` See [Cursor Rules docs](https://cursor.com/docs/context/rules). ### Windsurf Copy the \`skills/\` directory into your repo. Windsurf will pick up the Markdown files as context. You can also reference them from \`.windsurfrules\` at your project root. See [Windsurf Rules docs](https://docs.windsurf.com/windsurf/customize#rules). ### Other AI Tools For any AI coding tool: 1. Copy the \`skills/\` directory into your project 2. If the tool supports skills natively, it should discover the \`SKILL.md\` files automatically 3. Otherwise, point the tool's context or instructions configuration at the relevant \`SKILL.md\` files — each one references supporting files (\`reference.md\`, \`examples.md\`, \`rules/\`) for detailed guidance --- ## Automatic Detection (Claude Code only) The Claude Code plugin automatically detects Infrahub projects on session start by looking for: - \`.infrahub.yml\` or \`infrahub.toml\` configuration files - Schema files with \`version: "1.0"\` and \`nodes:\` or \`generics:\` keys When detected, Claude is instructed to use the appropriate skills for relevant tasks. ## Project Structure \`\`\` . ├── .claude-plugin/ │ └── plugin.json # Plugin manifest ├── hooks/ │ └── hooks.json # Hook definitions ├── hooks-handlers/ │ └── session-start.sh # Infrahub project detection script ├── skills/ │ ├── common/ # Shared references and rules │ │ ├── graphql-queries.md │ │ ├── infrahub-yml-reference.md │ │ └── rules/ │ ├── schema-creator/ # Schema design skill │ ├── object-creator/ # Data population skill │ ├── check-creator/ # Validation check skill │ ├── generator-creator/ # Generator automation skill │ ├── transform-creator/ # Data transform skill │ └── menu-creator/ # Navigation menu skill ├── CLAUDE.md # Project context ├── README.md # This file └── LICENSE # Apache 2.0 License \`\`\` ## Resources - [Infrahub Documentation](https://docs.infrahub.app/) - [Infrahub Schema Guide](https://docs.infrahub.app/guides/create-schema) - [Schema Library](https://github.com/opsmill/schema-library) - [OpsMill](https://opsmill.com/) ## License Apache License 2.0 - see [LICENSE](LICENSE) for details.