chat_with_codesnippet

GitHub Code Summarizer API

Overview

This project is a FastAPI-based service that leverages advanced natural language models and GitHub integration to summarize and explain code repositories. The Agent acts as a professional code reviewer capable of retrieving, analyzing, and providing insights into GitHub repositories based on user queries.

Features

• GitHub Integration: Interacts with GitHub repositories using access tokens for secure and efficient data retrieval.
• Customizable Queries: Accepts repository names and optional URLs for targeted analysis.
• Postgres Storage: Utilizes PostgreSQL for storing session histories and user interactions.
• Code Analysis: Reads, summarizes, and explains the purpose of code files in a repository.
• Query Flexibility: Can work with or without a specific repository URL.
• API Endpoint: Provides a /summarize_repo endpoint to handle user queries and return concise summaries.

Technologies Used

• FastAPI: For building and serving the REST API.
• Phi Framework: For creating a customizable and extendable agent.
• OpenAI GPT: A GPT model (gpt-4o-mini) for natural language understanding and generation.
• GitHub Tools: For interacting with GitHub repositories and retrieving relevant information.
• PostgreSQL: For storing session histories and agent interactions.
• dotenv: For managing environment variables.

Directory Structure

.
├── chat_with_code.py                     # Contains the API and agent logic
├── requirements.txt            # Python dependencies
├── .env                        # Environment variables (e.g., GitHub access token)

Installation

1. Clone the repository:

2. Set up a Python virtual environment:

   python3 -m venv venv
   source venv/bin/activate  # For Linux/macOS
   venv\Scripts\activate     # For Windows

3. Install dependencies:

   pip install -r requirements.txt

4. Set up environment variables: Create a .env file in the project root and add the following:

   GITHUB_ACCESS_TOKEN="your_github_access_token"

5. Set up PostgreSQL: Ensure a PostgreSQL database is running and accessible. Update the db_url in the PgAgentStorage configuration with your database credentials. you can use this command to run Pgvector in Docker from the documentation of Phidata:

      docker run -d \
     -e POSTGRES_DB=ai \
     -e POSTGRES_USER=ai \
     -e POSTGRES_PASSWORD=ai \
     -e PGDATA=/var/lib/postgresql/data/pgdata \
     -v pgvolume:/var/lib/postgresql/data \
     -p 5532:5432 \
     --name pgvector \
     phidata/pgvector:16

Usage

1. Run the FastAPI server:

   uvicorn main:app --host 0.0.0.0 --port 8000

   Example response:

   {
       "response": "The repository 'phidatahq/phidata' contains tools for data workflows..."
   }

Configuration

• GitHub Access Token: Ensure a valid GitHub personal access token is set in the .env file for repository interactions.
• Postgres Configuration: Update the db_url in the PgAgentStorage initialization to connect to your database.
• Agent Session Settings:
  • Modify session_id and user_id logic as required.
  • Adjust num_history_responses for the number of history records stored.

Dependencies

• fastapi
• phi
• langchain_openai
• uvicorn
• psycopg
• python-dotenv

Install all dependencies with:

pip install -r requirements.txt