Code Quality Intelligence Agent

Overview:

The Code Quality Intelligence Agent (CQIA) is a multi-agent system designed to automatically evaluate the quality of source code.
It combines static analysis, AST parsing, concurrent multi-agent evaluation, and a RAG-powered chatbot to provide deep insights into your project’s maintainability, security, and reliability.

---

🔍 Overview

CQIA leverages five specialized agents, each targeting a critical aspect of code quality:

1. Code Duplication Agent
   Detects redundant or repeated logic across files and modules, improving maintainability.

2. Security Agent
   Flags unsafe functions, insecure dependencies, and potential vulnerabilities.

3. Performance Agent
   Highlights inefficient loops, heavy memory usage, and resource bottlenecks.

4. Reliability & Fault Tolerance Agent
   Reviews exception handling and recovery strategies to ensure system resilience.

5. Complexity Agent
   Uses AST-driven structural analysis (cyclomatic complexity, nesting depth) to measure readability and maintainability.

---

⚙️ How It Works

1. Ingestion & Payload Creation

• Source files are ingested automatically.
• Each file is transformed into a payload containing:
  • Raw Code – original source for direct inspection.
  • AST (Abstract Syntax Tree) – structural representation of the code.
  • Metadata – file paths, function boundaries, and contextual info.

2. Concurrent Multi-Agent Analysis

• The payload is sent to all five agents concurrently, ensuring faster evaluations.
• Each agent produces issue reports specific to its domain.

3. RAG Chatbot

• A Retrieval-Augmented Generation (RAG) chatbot is integrated for interactive exploration.
• It leverages embeddings of raw code, AST fragments, and metadata to:
  • Describe project folder structures.
  • Summarize detected issues.
  • Provide natural language insights.

4. Report Generation

• All results are aggregated into a comprehensive quality report, combining duplication, security, performance, reliability, and complexity metrics.

---

📊 Output

• Structured JSON Reports for each agent.
• Consolidated Multi-Agent Report summarizing overall code quality.
• Interactive Chatbot to query project structure and issues in natural language.

Environment Variables

To run this project, you will need to add the following environment variables to your .env file

GROQ_API_KEY

Deployment

To deploy this project, clone the Repository and run the below commands

  pip install -r requirements.txt

  streamlit run app.py