add auto llm summary and tag sugestion to Videorama

[successbyfailure]

Use openai compatible endpoint to make summary and suggested tags.
Environment variables for summary prompt and tag suggest prompt and models.
Add a button to the video importer for "auto summary" and another for "auto tags". If transcription is generated save it on the video info for further reuse.

[successbyfailure]

@codex Work on this and make a PR

[chatgpt-codex-connector]

Summary

• Added OpenAI-compatible configuration, prompt helpers, and auto-summary/tag endpoints that reuse or generate transcriptions when available. videorama/main.pyL30-L145 videorama/main.pyL699-L727
• Allowed import submissions to merge client-provided metadata (such as generated transcripts) into stored entries for reuse. videorama/main.pyL567-L824
• Updated the import UI with “Auto resumen” and “Auto etiquetas” actions that call the new endpoints, track metadata, and surface assistant status to editors. templates/import_manager.htmlL329-L648
• Documented new LLM-related environment variables for model selection and prompt customization. example.envL15-L27

Testing

• ✅ python -m compileall videorama

View task →