entity-memory-management

What problem does it solve?

Entity Memory Management prevents your agent from forgetting important details (campaigns, metrics, decisions, pain points, and preferences) by extracting, persisting, and retrieving them as structured knowledge across sessions.

Core Features & Use Cases

• Entity extraction with a fixed taxonomy: Captures named entities and classifies them into campaign, metric, decision, pain_point, and preference schemas.
• Importance scoring for retrieval relevance: Assigns importance at extraction time so critical updates and preferences surface first later.
• Persistent storage and contextual recall: Upserts entities into a KV store with an index, then retrieves and injects only relevant entities for the current task.
• Decision and audit-friendly versioning: Updates existing entities instead of duplicating them, enabling change tracking through version increments.

Quick Start

Use the entity-memory-management skill to persist what your user says about campaigns, metrics, decisions, pain points, and preferences so the agent can reuse that context in the next session.