maybe completely vibe coded improvements

[hackz00]

What This Does

Added a bunch of improvements to make the tool easier to use and easier to debug.

Changes

New Features:

• Progress bar that shows how far along the investigation is and how much time is left
• You can press Enter to submit queries instead of clicking the button
• Shows what data is being sent to the LLM APIs in an expandable section
• Better logging - everything gets logged to robin.log and you can view logs in the UI
• Ollama models are automatically detected when you install them (no manual config needed)
• Added a refresh button to manually refresh the model list

UI Improvements:

• Cleaned up the settings section in the sidebar
• Better visual feedback when things happen
• Logs can be viewed in real-time with a checkbox

Under the Hood:

• Added logging throughout the codebase so we can see what's happening
• Better error handling for Ollama model detection
• Tracks timing for each stage of the investigation

Files Changed

• logger_utils.py (new file)
• ui.py
• llm.py
• main.py
• search.py
• scrape.py
• llm_utils.py

Testing

Everything should work the same as before, just with more features. Try running an investigation and you'll see the progress bar, or check the logs section in the sidebar.

No breaking changes - all existing functionality still works.

[apurvsinghgautam]

Thanks for your contribution, but the codebase has evolved significantly since this PR was opened, so I’m closing it. I’m also not prioritizing vibe-coded pull requests at the moment.