Cognifix

Cognifix is a command-line debugging agent that automatically detects programming errors, applies corrective fixes, and re-executes programs to validate successful execution.

The tool introduces a self-correcting execution pipeline that reduces the time developers spend identifying and fixing repetitive runtime errors.

---

Overview

Debugging is one of the most time-consuming tasks in software development. Developers often encounter common issues such as:

• Syntax errors
• Runtime exceptions
• Undefined variables
• Logical mistakes
• Incorrect calculations

Cognifix automates this process by executing code, capturing failures, and applying automated fixes until the program runs successfully.

---

Key Features

Automatic Language Detection

Cognifix detects the programming language of the provided source file.

Execution Engine

Each language is executed through a dedicated runtime engine.

QuickFix Mode

Applies immediate fixes for common errors to restore execution quickly.

DeepFix Mode

Performs deeper analysis of runtime errors and applies structured fixes.

Self-Correcting Execution Loop

After each fix, Cognifix automatically re-runs the program until it executes successfully or reaches the retry limit.

Modular Architecture

The system is designed to support additional languages and debugging strategies with minimal changes.

---

Architecture

Cognifix follows a modular debugging pipeline.

Source Code
    │
    ▼
Language Detector
    │
    ▼
Execution Engine
    │
    ▼
Error Capture
    │
    ▼
Repair Strategy
(QuickFix / DeepFix)
    │
    ▼
Re-Execution

Each component is independent, making the system easier to extend and maintain.

---

Project Structure

cognifix
│
├── main.py
│   CLI entry point
│
├── agent.py
│   Core debugging pipeline
│
├── language_detector.py
│   Language identification logic
│
├── engines
│   ├── python_engine.py
│   ├── javascript_engine.py
│   └── c_engine.py
│
├── modes
│   ├── quickfix.py
│   └── deepfix.py
│
├── test_programs
│   ├── test_code.py
│   ├── calculator_program.py
│   ├── grading_system.py
│   └── pi_program.py
│
└── README.md

---

Installation

Clone the repository.

git clone https://github.com/Tharshini272006/cognifix.git
cd cognifix

Ensure the following runtimes are installed:

• Python
• Node.js (for JavaScript execution)
• GCC (for C programs)

---

Usage

Run Cognifix from the command line.

QuickFix Mode

python main.py test_programs/test_code.py --mode quickfix

DeepFix Mode

python main.py test_programs/test_code.py --mode deepfix

---

Example Execution

Starting Cognifix...

Detected language: python

Error detected:
ZeroDivisionError: division by zero

Applying QuickFix...

Program executed successfully

---

Design Principles

Modularity

Execution engines and repair strategies are implemented as separate modules.

Extensibility

New languages and debugging strategies can be integrated without changing the core pipeline.

Fault Tolerance

Programs are executed multiple times until a fix succeeds or the retry limit is reached.

---

Extending Cognifix

Adding a New Language

1. Create a new engine inside engines/
2. Implement a runner function
3. Register the engine inside agent.py

Example:

engines/java_engine.py

---

Adding a New Repair Strategy

Create a new debugging strategy inside the modes/ directory and integrate it into the debugging loop.

---

Future Improvements

• AI-assisted debugging using LLMs
• IDE integrations (VS Code extension)
• Static code analysis
• GitHub pull request auto-repair bot
• Support for additional programming languages

---

Author

Tharshini GitHub: https://github.com/Tharshini272006

---

License

This project is released under the MIT License.