What problem does it solve? AI agents lose all context between sessions, forcing users to repeat preferences, configurations, and decisions. This Skill evaluates each conversation turn and persists only durable, actionable facts to mem0 long-term memory so future sessions start with relevant context. ## Core Features & Use Cases - Four-Gate Fact Triage: Every candidate fact must pass future-utility, novelty, factual, and safety gates before storage, preventing memory pollution from small talk, tool output, and transient state. - Categorized Memory Operations: Store, search, update, consolidate, and delete memories across eight categories (identity, preference, decision, rule, project, configuration, technical, relationship) with per-category retention policies. - Credential Safety: Never stores secrets or tokens; records only that a credential was configured, with pattern-based detection of API keys and auth tokens. - Use Case: A user mentions they switched monitoring from Datadog to Grafana due to cost. The Skill stores a temporally anchored decision memory, and weeks later a new agent session recalls this context when the user asks about their observability stack. ## Quick Start Ask the agent to remember that you prefer terse responses with no trailing summaries, and it will store that preference for all future sessions.
