@divinci-ai/types

TypeScript type definitions for the Divinci AI SDK

Installation

npm install @divinci-ai/types

Note: This package is automatically installed as a dependency of @divinci-ai/client, @divinci-ai/server, and @divinci-ai/mcp. You typically don't need to install it directly.

Usage

import type {
  Message,
  ChatResponse,
  Workspace,
  Release,
  RagDocument,
} from "@divinci-ai/types";

// Use types in your code
const message: Message = {
  role: "user",
  content: "Hello!",
};

Available Types

Chat Types

• Message - Chat message structure
• ChatResponse - Response from chat API
• StreamChunk - Streaming response chunk

Workspace Types

• Workspace - Workspace configuration
• WorkspaceSettings - Workspace settings

Release Types

• Release - Published release
• ReleaseConfig - Release configuration

RAG Types

• RagDocument - RAG document
• RagVector - Vector store configuration
• SearchResult - Search result

API Types

• ApiKey - API key structure
• ApiError - Error response

Zod Schemas

This package also exports Zod schemas for runtime validation:

import { MessageSchema, ChatResponseSchema } from "@divinci-ai/types";

// Validate at runtime
const result = MessageSchema.safeParse(unknownData);
if (result.success) {
  const message = result.data;
}

Documentation

Full documentation: sdk.divinci.ai/reference/types

License

MIT - see LICENSE for details.