What problem does it solve? Engineers often lose valuable technical insights gained during coding sessions and PR reviews because they never write them down. This Skill captures those learnings and appends them in a structured format to your Obsidian Daily note, building a searchable personal knowledge base over time. ## Core Features & Use Cases - Conversation-Based Extraction: Analyzes the current session to identify what you asked, what you learned, and unresolved questions, with emphasis on Go, databases, infrastructure, and architecture topics. - PR Review Extraction: Uses the GitHub CLI (gh) to fetch PRs you reviewed or authored today, reads diffs and review comments, and distills design intent and review feedback into learnings. - Structured Markdown Output: Appends entries with code blocks, Before/After comparisons, and "why" explanations to ~/obsidian/Daily/{YYYY-MM-DD} {曜日}.md, skipping already-recorded PRs. - Use Case: After a day of reviewing backend PRs, run the review mode to automatically document the design patterns and review feedback you encountered, complete with code excerpts. ## Quick Start Ask the AI to record today's learnings from this conversation into your Obsidian daily note, or request it to also include learnings from today's GitHub PR reviews.