Summary

Frank hosts a two‑hour live coding session improving his open‑source Notebookmark app by adding AI-powered reading suggestions via a new ResearchService (using Reka.ai) and building a Blazor search page and grid, while also demoing a custom documentation MCP server for targeted docs lookup. Along the way he discusses dev environment quirks (Debian/.NET version clash, switching to an Arch-based Omarchy setup), tweaks the UI (navigation, wider form fields), pushes code, and chats with viewers amid Halloween vibes. The stream ends with progress committed and a short roadmap to add prompt configurability, domain allow/block filters, and structured results for the suggestions list.

📺 - Twitch archive - stream no.270

Replay


Project

All the code for this project is available on GitHub: notebookmark - https://github.com/FBoucher/notebookmark

Game Results

Statistics

Notes/ References / Snippets

  • https://www.csadvent.christmas/
  • https://reka.ai/
  • Awesome Java guy: https://github.com/bbenz