- [x] API final plan -- separate / openrouter, is there difference (gemini, claude), which model to use cheapest & best - [x] try achieving auto-context myself - [x] Make LLM be able to read title of note as well as content - [ ] figure out how embedding and vault QA works - [x] custom chat type based system prompt q5twfonhuRjzgEaoVAxlDit5jWrCgaOjvDBFRMOBc ## Custom function achieved - css tweak -- increasing paragraph distance - Current note auto context & context switch - per chat System Prompt - [ ] youtube video transcription --- Copilot -- how to / should I catch up with the progress of the official branch gpt5 parameter implement my own version of "Proejct Mode" in copilot? (separate context, system prompt) my own good version of vaultQA?