Skip to content

Latest commit

 

History

7 Commits

Folders and files

NameName
Last commit message
Last commit date
 
 
 
 
 
 
 
 
 
 

Repository files navigation

TestPilot AI — PS-10

Self-Improving Test Generation Agent for the 6-hour Agentic AI hackathon.

Workflow

User Code → Code Analyzer Agent → Test Generator Agent → PyTest Runner → Failure Analyzer/Test Improver → Re-run Tests

Setup

python -m venv venv
venv\Scripts\activate
pip install -r requirements.txt

Copy .env.example to .env and add your Groq API key:

GROQ_API_KEY=your_api_key_here

Run:

streamlit run app.py

Project Diagram :

             👨‍💻 DEVELOPER
                   │
                   │ Python Code
                   ▼
          ┌─────────────────┐
          │ 🧠 CODE ANALYZER│
          └────────┬────────┘
                   │
             Testing Plan
                   ▼
          ┌─────────────────┐
          │ 🧪 TEST         │
          │    GENERATOR    │
          └────────┬────────┘
                   │
              PyTest Code
                   ▼
          ┌─────────────────┐
          │ ▶️ TEST RUNNER  │
          └────────┬────────┘
                   │
              Test Result
              /          \
           PASS            FAIL
            │                │
            │                ▼
            │       🔧 IMPROVEMENT
            │          AGENT
            │                │
            │                ▼
            │          Improved Test
            │                │
            └───────◄────────┘

Why not Claude or GPT?

Factor Groq OpenAI GPT Claude
API integration Simple Simple Simple
Fast inference Very strong Strong Strong
Code generation Good Good/very strong depending on model Good/very strong depending on model
Good for hackathon demo ✅ ✅ ✅
Requires their respective API key Yes Yes Yes
Our current code ✅ ❌ ❌

Demo

Paste Python code, then click Generate & Run Tests. The app analyzes the code, generates pytest tests, executes them, and if they fail, asks the AI to improve the tests and runs them again.

This is a controlled hackathon prototype. Do not execute arbitrary untrusted code with this local subprocess runner in production.

👨‍💻 About the Author

Hi, I'm Ronit Raj, a Computer Science student and aspiring Software & AI Developer passionate about building practical projects and exploring Agentic AI, Generative AI, Web Development, and Data Structures & Algorithms.

I enjoy turning ideas into real-world applications, experimenting with new technologies, and continuously improving my problem-solving and development skills.

🚀 What I Work With

  • 🐍 Python | C++ | Java | JavaScript
  • 🤖 Generative AI | Agentic AI | LLM Applications
  • 🌐 HTML | CSS | React | Tailwind CSS
  • 🧠 Data Structures & Algorithms
  • 🗄️ SQL | MySQL | MongoDB
  • 🛠️ Git | GitHub | Linux | FastAPI

🔗 Connect With Me

💡 Building, learning, and improving — one project at a time.

About

Self-Improving Test Generation Agent using Agentic AI, Groq and PyTest

Resources

Stars

1 star

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages