We Published a Royalti.io API Skill for AI Coding Assistants

Every integration with the Royalti API starts the same way: you paste docs into your AI assistant, explain the auth flow, then re-explain it next session. The Royalti.io API Skill eliminates that. Install it once, and your assistant knows every endpoint, authentication pattern, and webhook event before you ask your first question.

Get the skill on GitHub

What Is an Agent Skill?

An Agent Skill is an open standard for teaching AI assistants domain-specific knowledge. Instead of pasting API docs into every conversation, you install a skill once, and your assistant loads it automatically when relevant.

In December 2025, Anthropic published the Agent Skills specification, which uses a directory containing a SKILL.md file with metadata, optional scripts, templates, and reference docs. So that when you make a request that matches a skill's domain, the assistant loads only what it needs. This design is called "progressive disclosure."

The specification was originally developed by Anthropic and has since been adopted by 25+ AI tools.

Works Everywhere

The skill is compatible with every major AI coding assistant that supports the Agent Skills standard:

• Claude Code: Native skill support via .claude/skills/
• Claude Desktop: Upload as a zip via Settings > Capabilities > Skills
• Cursor: Copy to .cursor/skills/
• OpenAI Codex: Copy to .codex/skills/
• GitHub Copilot (VS Code): Copy to .github/skills/
• OpenClaw: Copy to ~/.openclaw/skills/
• Gemini CLI: Copy to .gemini/skills/
• Amp: Copy to .amp/skills/
• Goose: Copy to .goose/skills/
• Roo Code: Copy to .roo/skills/
• OpenCode: Copy to .opencode/skills/
• Letta: Copy to .letta/skills/

Or you could use a universal installer:

1# OpenSkills
2npx openskills install Royalti-io/royalti-api-skill
3
4# Vercel Skills
5npx skills add Royalti-io/royalti-api-skill
Copy

What the Skill Covers

Once installed, the rest is easy. Your AI assistant knows the full Royalti REST API.

Authentication: The two-step JWT login flow, workspace and user API keys (RWAK/RUAK), OAuth providers, rate limits, and RBAC roles. Ask your assistant to "authenticate with the Royalti API" and it writes the correct code immediately.

Every Endpoint (CRUD patterns for all resources): users, artists, assets, products, splits, royalties, accounting, payments, files, labels, releases, expenses, revenue, notifications, downloads, and DDEX delivery (the industry-standard format for sending releases to streaming services). Ask your assistant to "list all artists with pagination" and it knows the exact query parameters.

Webhooks: Event types across five categories. Ask your assistant to "set up a webhook listener for payment events" and it generates the configuration, event types, and HMAC signature validation code.

WebSocket Events: Real-time file processing updates via Socket.io. Ask your assistant to "track royalty file upload progress" and it writes the connection code with the correct event names.

Integration Patterns: Step-by-step flows for common scenarios like catalog sync, royalty file processing, and split management. Ask the assistant to "sync my catalog to Royalti" and it generates the full sequence of API calls in the right order.

Quick Install

One-liner (any agent):

1npx openskills install Royalti-io/royalti-api-skill
Copy

Auto-updating (symlink):

1git clone https://github.com/Royalti-io/royalti-api-skill.git
2ln -s "$(pwd)/royalti-api-skill/skills/royalti-api" .claude/skills/royalti-api
Copy

Then run git pull inside the clone whenever you want the latest version. The README has paths for every supported tool.

Why an Open Skill?

Most platforms keep their API knowledge locked inside proprietary docs that AI assistants can't reach. Every new conversation starts from scratch.

Publishing the Royalti API as an installable skill means:

• Zero context-setting: That your assistant already knows the API before you ask your first question
• Cross-platform: The same skill works in whichever tool you prefer
• Always available: There is no need to copy-paste docs or link to reference pages
• Community-maintained: It is Open source, MIT licensed, and contributions are welcome.

What's Next

This skill is a reference companion to the Royalti MCP Server, our existing integration that lets AI assistants interact with your Royalti workspace directly. If you haven't set it up yet, the MCP setup guide walks you through it.

The skill already includes language-specific code examples for JavaScript, Python, and PHP, covering authentication, CRUD operations, file upload, webhooks, and WebSocket events.

Next up: specialized workflow skills for royalty file processing and DDEX delivery.

Questions or suggestions? Open an issue on the repo.

Sources

• Agent Skills Specification: The open standard for AI agent capabilities
• Skills.sh Directory: Browse and discover Agent Skills packages
• OpenSkills CLI: Universal skills installer for AI coding agents
• Vercel Skills CLI: Vercel's open agent skills tool
• Anthropic Agent Skills Announcement: VentureBeat coverage