Radzor Components

The official registry of components for the Radzor Platform.

What are Radzor Components?

Radzor components are fully typed, framework-agnostic building blocks for backend logic, APIs, and AI workflows. They are not distributed as node_modules packages. Instead, you download the source code directly into your repository.

Preventing LLM Hallucinations

Every component in this repository includes:

1. The source code (TypeScript, Python, etc.)
2. A radzor.manifest.json compliant with the RCS Specification
3. An llm/integration.md guide

This ensures that when you ask an AI agent to use the component, it reads the manifest, understands the exact inputs/outputs, required API keys, and events, and generates accurate integration code without relying on outdated training data.

The Catalog

The registry contains components across categories:

• AI: llm-completion, speech-to-text, text-to-speech, embeddings-store, vector-search...
• Audio/Media: audio-capture, image-processor, video-transcoder...
• Auth: auth-oauth, api-key-auth, jwt-verify...
• Payment: stripe-checkout, webhook-receiver...
• Storage/DB: file-upload, sql-query, redis-cache...

Explore the full catalog on radzor.io/components.

Usage

Use the Radzor CLI to install a component:

npx radzor@latest add speech-to-text llm-completion

Or scaffold a complete AI Workflow (Recipe):

npx radzor@latest recipe add voice-bot

Contributing

We welcome community contributions to build a standard registry of LLM-friendly components.

To add a new component to the registry:

1. Fork this repository.
2. Scaffold your component using the CLI:

   npx radzor@latest create @radzor/my-component -c ai

3. Implement the logic in src/index.ts and write the LLM instructions in llm/integration.md.
4. Run npx radzor validate . inside your component folder to ensure your manifest passes the RCS Schema validation.
5. Open a Pull Request.

Links

• Radzor Platform
• RCS Specification
• Radzor CLI

License

MIT