martiendejong 

Core data models and contracts for the Hazina ecosystem. Provides chat message models, LLM response wrappers with token usage tracking, tool definitions, image generation models, and shared interfaces used... More information

Provider-agnostic interface for LLM interactions. Defines the ILLMClient interface with support for chat completions, streaming, structured JSON responses, image generation, text-to-speech, embeddings, and tool... More information

Utility functions for document and token processing. Includes TokenCounter for GPT token counting, DocumentSplitter for chunking by token limits, PartialJsonParser for streaming JSON, and helpers for checksums,... More information

Shared models and DTOs for DevGPT tools

Domain models for DevGPT content generation tools

Core functionality for DevGPT generation tools

Document-augmented LLM response orchestration for DevGPT. Provides IDocumentGenerator for composing RAG context with prompts, streaming responses, and safe file modification handling. Core component for... More information

Data access layer for DevGPT generation tools

High-level agent factory for building autonomous AI agents in DevGPT. Provides configuration parsing, built-in tool collections (file ops, git, dotnet, npm, BigQuery, email, WordPress), and multi-agent flow... More information

Tool calling extensions for DevGPT LLM clients. Provides reusable tools that LLMs can invoke including Claude CLI execution, web page scraping, and tool context base classes. Enables function calling and... More information

Multi-provider abstraction layer for Hazina AI Framework - Seamless LLM provider switching with failover, health monitoring, and cost tracking

Package Description

Text extraction utilities for PDF, Word, Excel and other document formats

Core data models and contracts for the DevGPT ecosystem. Provides chat message models, LLM response wrappers with token usage tracking, tool definitions, image generation models, and shared interfaces used... More information

Utility functions for document and token processing. Includes TokenCounter for GPT token counting, DocumentSplitter for chunking by token limits, PartialJsonParser for streaming JSON, and helpers for checksums,... More information

Provider-agnostic interface for LLM interactions. Defines the ILLMClient interface with support for chat completions, streaming, structured JSON responses, image generation, text-to-speech, embeddings, and tool... More information

Package Description

Vector embedding storage for semantic search in Hazina. Provides IEmbeddingStore interface with PostgreSQL/pgvector backend, batch operations, similarity matching, and embedding generation service. Essential... More information

Document storage and retrieval system for RAG (Retrieval-Augmented Generation) in Hazina. Provides IDocumentStore interface with support for text/binary documents, chunking, metadata management, and relevancy... More information

Package Description