Enhanced Aria v2 with LangChain Agent Integration

[adrew25]

Overview

This pull request significantly enhances Aria by integrating LangChain, enabling advanced agent capabilities such as tool usage, dynamic decision-making, and intelligent responses. This marks the transition of Aria from a basic conversational AI to a capable assistant with contextual understanding and tool-based reasoning.
New Features
LangChain Agent Integration

Tool Usage: Automatically identifies and uses the appropriate tools based on the user’s input.

Contextual Understanding: Interprets user intent and context to trigger relevant agent actions.

Modular Integration: Seamlessly connects with existing components like voice interface, text-to-speech (TTS), and UI.

Personality-Aware Responses: Combines tool outputs with Aria’s unique sarcastic personality in generated responses.

Agent Logic

Intent Recognition: Parses user queries to detect goals and desired outcomes.

Tool Selection: Selects and executes the most appropriate tool in real-time.

Result Integration: Merges results into coherent responses matching the assistant’s tone and style.

Graceful Degradation: Falls back to default messaging or basic responses when tools are unavailable or fail.

Dependencies

requirements.txt has been updated to include the necessary packages:

langchain>=0.1.0
langchain-community>=0.0.20
duckduckgo-search>=4.0.0
ddgs>=1.0.0
wikipedia>=1.4.0
requests>=2.31.0