What problem does it solve? It turns scattered social media traces, chat logs, and posting history into evidence-based self-knowledge, producing structured JSON profiles and readable Markdown reports instead of vague personality guesses. ## Core Features & Use Cases - Five archaeology modes: past-life habit tracing, cringe archaeology of old posts, public-vs-private persona comparison, digital legacy audit with time-cost conversion, and five-style epitaph generation. - Versioned profiles: every analysis writes a JSON profile plus Markdown report, with snapshot, rollback, correction, and append workflows managed by profile-manager.py. - Privacy-first data intake: supports user narration, exported chat files, screenshots, and browser-assisted capture of the user's own logged-in pages, with all analysis kept local. - Use Case: A user exports years of WeChat and Weibo history, runs the legacy audit mode, and receives a report converting 2847 hours of screen time into books, trips, and a concrete existential follow-up question. ## Quick Start Say "我想做遗产清算" or "analyze my digital life" and answer the agent's two or three intake questions to generate your first profile.