What problem does it solve? It turns scattered workout notes into a persistent, local-file training system where an AI coach tracks weights, reps, RPE, recovery, and long-term progression across sessions, with an optional RPG layer that converts training results into levels, attributes, and narrative advancement. ## Core Features & Use Cases - Workout logging and state tracking: Appends daily sessions to user-data/sessions/YYYY-MM-DD.json and recomputes EXP, level, attributes, recovery, and personal bests in CURRENT-STATE.json via zero-dependency Python scripts. - Recovery-aware planning: Adjusts recommended intensity based on days since last session plus six correction factors such as sleep, pain, and recent RPE trends, including comeback protocols after long breaks. - Optional RPG story mode: Maps push/pull/legs training to narrative scenes using a reusable story engine, narrative templates, and a ready-made default world, all customizable per user. - Use Case: After a two-week break you tell the AI "开始训练"; it reads your recent sessions, detects the interruption, reduces your bench press load by 20-30%, guides the session, records the results, updates your level, and advances your story chapter. ## Quick Start Tell the AI "开始训练" or "初始化档案" and answer the four setup questions about your goal, equipment, injury history, and whether you want to train today.
