Interface IAutoBeProps<Model>

Compiler factory function for TypeScript, Prisma, and OpenAPI operations.

Factory function that creates compiler instances required for the sophisticated AST-based development pipeline. The compiler handles validation, transformation, and code generation across all development phases including Prisma schema compilation, OpenAPI document validation, and TypeScript code compilation.

For high-performance scenarios with multiple concurrent users, the compiler can be separated into dedicated worker processes to prevent blocking the main agent during computationally intensive compilation operations.

Optional behavioral configuration for localization and context.

Customizes the agent's communication style, language preferences, and geographical context to provide personalized vibe coding experiences. Configuration includes locale settings for internationalized responses and timezone information for temporal context awareness.

These settings influence how the agent communicates with users, interprets regional requirements (such as regulatory considerations), and handles time-sensitive operations throughout the development process.

Optional conversation and development histories for session continuation.

Enables resuming previous vibe coding sessions by providing the chronological record of past conversations, development activities, and generated artifacts. When provided, the agent reconstructs its internal state from these histories, allowing seamless continuation of development work.

This capability supports iterative development workflows where users can return to modify, enhance, or extend previously generated applications while maintaining full context of earlier decisions and implementations.

AI model type specification for type-safe function calling schema generation.

Determines the specific AI model schema used for generating function calling interfaces through typia.llm.application(). This type parameter ensures compile-time type safety and enables model-specific optimizations in the AI function calling interface generation process.

Common values include "chatgpt" for OpenAI models, "claude" for Anthropic models, "deepseek" for DeepSeek models, and "llama" for Meta Llama models. The choice affects function calling capabilities, parameter limitations, and schema requirements throughout the vibe coding pipeline.

Note that Google Gemini ("gemini") is not supported due to its lack of reference types and union types support required for OpenAPI document composition in the vibe coding process.

Token usage information.

You can start token usage tracing by assigning this property.

If you assign IAutoBeTokenUsageJson value, the token usage tracing would be from the value. Otherwise you assign the AutoBeTokenUsage typed instance, the tracing would be binded to the instance.

AI vendor configuration for service provider integration.

Defines the complete AI service connection including the OpenAI SDK instance, model identifier, request options, and concurrency controls. This configuration enables the AutoBeAgent to connect with various AI providers while maintaining consistent functionality across the entire automated development workflow.

The vendor settings determine the AI capabilities available for requirements analysis, database design, API specification, testing, and implementation phases of the vibe coding process.