What problem does it solve? Turning months of raw development history into a coherent story is hard when observations are scattered across hundreds of sessions. This Skill fetches the complete claude-mem timeline for a project and produces a structured narrative report covering architecture evolution, breakthroughs, debugging sagas, and token economics. ## Core Features & Use Cases - Full Timeline Retrieval: Fetches every observation, session boundary, and summary from the claude-mem worker API, with automatic git worktree detection to resolve the correct parent project. - Subagent-Based Analysis: Delegates the timeline to an analysis agent that writes a 3,000-6,000 word "Journey Into [Project]" report with ten required sections including token economics and memory ROI. - SQLite Metrics Queries: Runs SQL against the claude-mem database to quantify discovery tokens, compression ratios, recall savings, and the highest-value observations. - Use Case: Ask for a journey report on a long-running project to understand how its architecture evolved, which bugs consumed the most effort, and what a new developer should learn from its history. ## Quick Start Ask the assistant to write a timeline report for your project, for example by saying "Write a journey report for the tokyo project" while the claude-mem worker is running.
