Context Retention System

Overview

This project is a comprehensive system designed to aid software developers in generating and debugging code efficiently using context retention and AI-driven language models. The system integrates capabilities such as extracting function/method details, refining user prompts, and interacting with large language models (LLMs) through Groq's API.

---

Key Features

1. Centralized Data Extraction

• Scrapes function and method details such as:
  • Input/Output Parameters
  • Docstrings
  • Dependencies
  • File Paths
• All extracted details are stored in a centralized JSON file for efficient access and future usage.

2. User-Defined Prompts

• Allows users to define an initial prompt that specifies:
  • Objective: The task or goal the user wants to achieve.
  • Dependencies: Relevant methods or functions, mentioned in double quotes (e.g., "generate" method of "GenerateGoal" class).

3. Prompt Refinement

• The initial prompt is processed by the PromptGenerator class, which:
  • Adds detailed information about the methods/functions referenced in the user’s prompt (e.g., docstrings, parameter details).
  • Refines the prompt to provide complete context for the LLM.

4. LLM Interaction via Groq

• The refined prompt is passed to the call_llm method, which interacts with Groq's API to query openly available LLMs.
• Supported models can be found on Groq's Model Page.

5. End-to-End Flow

• The entire process, from prompt creation to result inspection, is demonstrated in the exp.ipynb notebook.
• Users can test the system end-to-end and verify results.

---

Folder Structure

• function_extrcator_v2.py # Script for extracting function/method details
• prompt_generator.py # Class to refine user-defined prompts
• call_llm.py # Function to interact with LLM via Groq
• exp.ipynb # End-to-end example notebook
• README.md # Project documentation
• requirements.txt # Dependencies