DebuggAI (debugg-ai-py)

Debugg AI's Python sdk for enabling your personal AI QA engineer

✨ Why DebuggAI?

Most AI coding tools focus on writing code. DebuggAI focuses on the other 50 % of an engineer’s life: getting it to run.

• AI Test Suites — We let you focus on the code while our QA engineering agent handles the rest. DebuggAI builds & runs test suites in the background to ensure old code continues to run and new code avoids possible edge cases BEFORE it gets to a PR, or worse to your users.
• 1‑line monitoring SDK — drop‑in client (Node, Python, Go) that captures rich runtime context remotely similar to Sentry or Datadog
• AI Debug — Errors are instantly sent to failure lines in your IDE so you can see what happened and why, making solving it easy.
• Instant Fix Suggestions — one‑click patches and PRs generated from stack‑trace + context
• Source‑map de‑minification — readable traces even for bundled / minified front‑end code
• Branch‑aware log search — slice errors by branch, release, or feature flag to zero in fast

📺 Demo - Get Instant Insight Into Runtime Issues

🔍 Typical workflows:

1. You use your favorite AI agent to write code
2. You run your app and it crashes (ah whyyyyy!)
3. DebuggAI sees the error, grabs the full stack trace + context, and uses it to generate a solution & show you EXACTLY where to look
4. You review the solution, edit it locally if needed, and apply it

🔍 How it works

🖥️ Core IDE Features

🚀 Getting Started

1. Install the extension

   • VS Code Marketplace
   • Jetbrains coming soon

2. Create a project

   • Sign up & create a project in the DebuggAI app

     

3. Add the Python Logging SDK (using pip)

     pip install debugg-ai-sdk
   

4. Initialize (one line):

   • Get the initialization code from the DebuggAI app

     

   • Initialize the SDK

       # app.py
       import debugg_ai_sdk
       debugg_ai_sdk.init(
        dsn=f"<your_project_dsn>",
        # Depends on your implementation but info level provides 
        # more helpful context to the agents than ERROR-only.
        level='info',
        environment="local",
        # Readable name to differentiate local computers for engineers
        host_name="tj-home-desktop",
        # other optional params..
       )
     

   • Log errors

       class TestClass:
         def __init__(self):
             
             self.test_var = "test"
             
         def divide_by_zero(self):
             return 1 / 0
     
         def test_function():
             # Fundtion that pretends to do something
             logger.info("Doing something")
             
             test_class = TestClass()
             test_class.divide_by_zero()
             
     
       def main():
           logger.info("Hello, world!")
     
           test_function()
     
       if __name__ == "__main__":
           main()
           
     

5. Trigger an error – head back to the IDE and watch DebuggAI suggest a fix ⚡

Full walkthrough ▶ docs.debugg.ai/getting-started

🛠️ Configuration

You can log in to your DebuggAI account directly in the extension, and then it will automatically connect to your project.

Contact & Support

If you have any questions or need personalized support:

• Email: support@debugg.ai
• Discord: Join our Discord community at DebuggAI Discord Server
• Documentation: Official DebuggAI Docs

🤝  Interested in Contributing?

We're looking to expand the DebuggAI team!

If you're interested in joining the team or contributing to the project, please reach out to us at hello@debugg.ai.

📜 License & Credits

• Code: MIT © 2025 Debugg, Inc.
• Foundation: proudly built on open-source technology.

Attribution

We at Debugg AI want to thank the open-source community for their contributions. Particularly Sentry for the work on this SDK. DebuggAI is building the first fully AI QA Engineer that can automatically generate test suites and highlight issues in your app, but Sentry continues to be a great option for Application Monitoring. Use both for the best results!