revise evaluation to be generic

revise evaluation prompt

support updating evaluation prompt in UI

refine evaluation and fix CSV download problem

support evaluating coach direct prompt

support conversation evaluation

tune performance

fix scripts for new prompts

feat: Add database-backed prompt management system with repository pattern

refactor: Adopt OpenAI Responses API for conversation continuity and improve test reliability

feat: Add comprehensive Basic Auth and enhance prompts with developmental theory

feat: Add auto-generated conversation titles with inline editing