AGENTS.md

Project Agents.md Guide for OpenAI Codex

This Agents.md file provides guidance for OpenAI Codex and other AI agents working with this codebase.

Project Structure for OpenAI Codex Navigation

• /pixaris: Core library code
  • /data_loaders: dataset loaders
  • /experiment_handlers: experiment orchestration
  • /feedback_handlers: feedback handling
  • /generation: image generation modules
  • /metrics: evaluate experiment
  • /frontend: Gradio UI components
  • /orchestration: glue between the components
  • /utils: helper utilities
• /examples: Example scripts demonstrating usage
• /test: Pytest suite for maintaining quality
• /docs: Documentation built with Sphinx
• /assets: Static images and other assets; these should not be modified directly

Coding Conventions for OpenAI Codex

General Conventions for Agents.md Implementation

• Use Python with type hints for all new code
• Follow the existing code style
• Provide descriptive variable and function names
• Add docstrings for non-trivial logic
• Format code with make fmt and lint with make lint

Gradio UI Components Guidelines for OpenAI Codex

• Use Gradio Blocks when adding UI functionality
• Keep components small and well typed
• Prefer a functional style with clear input and output annotations

Styling Standards for OpenAI Codex

• Gradio theming is handled centrally; avoid custom CSS unless necessary

Testing Requirements for OpenAI Codex

Run tests with:

make test

Pull Request Guidelines for OpenAI Codex

1. Provide a clear description of the changes
2. Reference related issues if applicable
3. Ensure make lint and make test succeed
4. Include screenshots for UI changes
5. Keep each PR focused on a single topic

Programmatic Checks for OpenAI Codex

Before submitting changes, run:

make lint
make test

All checks must pass or note any issues in the PR description.