memory-governance

What problem does it solve?

Memory governance addresses the challenge of maintaining high-quality, consistent memories across multiple layers (Instance, Template, Actant) in AI agents. It provides mechanisms to promote valuable memories to broader layers, apply decay to stale items, deduplicate duplicates, and condense learnings for reliable reuse.

Core Features & Use Cases

• Layered memory promotion: elevate important memories from Instance to Template or Actant layers for reuse.
• Deduplication and condensation: merge duplicates and produce concise, high-quality memories.
• Memory decay and governance: apply decay to low-confidence memories and maintain audit trails for changes.
• Use Case: Apply to long-running agents to keep knowledge up-to-date while avoiding memory bloat across sessions.

Quick Start

Promote valuable memories to higher layers, prune stale items, and consolidate duplicates for consistent recall.