I am starting to work on it .... Can you clarify testGPT’s role here? Is it a custom script, an external API, or part of Keploy’s existing tooling? Are there existing examples/docs I can reference?

I am starting to work on it .... Can you clarify testGPT’s role here? Is it a custom script, an external API, or part of Keploy’s existing tooling? Are there existing examples/docs I can reference?

Hi @Parijat-Ghosh , testGPT is the keploy's github action you can find here. It uses bash scripts to run and get triggered once it is in pipeline

Thanks a lot!

@Sonichigo will be picking this up as well!

Hey People! Over the weekend, I worked on a prototype for this issue, I call it PRism (get it? PR-ism? lol). I just wanted to quickly share what I have built so far:

1. LLM Integration:

• Designed the system to be LLM-agnostic, allowing integration with any open-source or closed-source LLM.
• The tool also recommends which LLM to use based on use-case. Note: This is when you self-host the tool.

2. Security Enhancements:

• I used trivy to setup the security dependency checks.

3. Static Analysis and Linting:

• Integrated Keploy to run test generation workflows.
• Added ESLint and GolangCI-Lint for comprehensive static code analysis.

4. Line-by-Line Suggestions (CodeRabbit-style):

• Implemented inline suggestions for coding style, security vulnerabilities, and best practices directly in pull requests (PRs).
• Ensures actionable feedback, improving developer experience.

5. Modular Webhook System:

• Built a webhook-based system to integrate seamlessly with CI/CD workflows and version control platforms (e.g., GitHub, GitLab).
• Configurable trigger points to optimize performance (e.g., only running on certain branches or PR events).

Demo Video :

Looking forward to your feedback on this!