What problem does it solve? Important facts shared in conversation are easily lost between sessions, and manually tracking user preferences or decisions leads to stale, contradictory information. This Skill captures durable facts from text and persists them into db0 memory with the correct scope so they remain available across sessions. ## Core Features & Use Cases - Fact Extraction: Analyzes provided text to identify durable facts worth remembering, such as preferences, decisions, and personal context. - Scope Assignment: Automatically assigns each fact to the appropriate scope — user (cross-session), agent (all sessions), or session (current context). - Superseding Corrections: Detects corrections to existing facts, searches for the old memory, and marks it as superseded so stale information is excluded from search while preserved for audit. - Use Case: When a user says "remember that I prefer dark mode" and later says "actually I switched to light mode", the Skill stores the new preference and supersedes the old one, keeping memory accurate. ## Quick Start Ask the assistant to remember a fact, for example: remember that I prefer TypeScript for all new projects.