What problem does it solve? Long AI coding sessions lose their context when they end — decisions, files touched, and next steps vanish, forcing the next session to start from nothing. This Skill wraps up the active session by writing a structured summary into its registered notes.md so work can be resumed later without re-discovery. ## Core Features & Use Cases - Session summary generation: Collects decisions made, files touched, open questions, and a prioritized next-steps queue from the current conversation and updates the matching sections in notes.md. - Frontmatter refresh: Auto-attaches GitHub PR links (via gh CLI and conversation scan) and ticket IDs, recording per-ticket status without overwriting existing values. - Session history marker: Appends a dated closing entry that marks the session Closed in the dashboard, with duplicate-close protection. - Optional knowledge distillation: When a vault is configured, promotes reusable decisions through the /learn skill, and can stage skill proposals under ~/.claude/skills-pending/ when enabled. - Use Case: After finishing a feature branch, run the close command to record what was decided, link the PR and ticket, and leave a clean next-steps queue for whoever resumes the session. ## Quick Start Ask the assistant to wrap up and save this session's notes by saying "close this session and save the session notes".