v0.1.0 — MCP-Based Document Intelligence Platform

This release introduces an enterprise-style document intelligence platform built on the Model Context Protocol (MCP), enabling Large Language Models (LLMs) to dynamically discover tools, retrieve documents, and generate grounded answers through protocol-driven orchestration.

✨ Key Capabilities
• Model Context Protocol (MCP) Server
• Exposes document search tools, PDF resources, and reusable prompt templates via standardized MCP descriptors.
• Persistent Vector Retrieval
• PDF ingestion and semantic indexing using ChromaDB and OpenAI embeddings for accurate, metadata-aware search.
• LLM Tool-Orchestrated Reasoning
• Allows the model to decide between direct responses and tool-assisted retrieval at runtime.
• Role-Aware Conversational Memory
• Structured handling of system, user, assistant, and tool roles to ensure OpenAI-compliant tool-calling flows.
• Async MCP Client (STDIO Transport)
• Client automatically spawns and manages the MCP server, executes tools, and maintains conversation state.
• Interactive CLI Interface
• Supports live queries, document exploration, prompt execution, and debug tracing.