memory-systems - Design persistent memory for agents
Designs persistent agent memory across sessions, including entity tracking, temporal validity, retrieval strategies, consolidation, and benchmark selection.
Tags
Updated: 2026-10-02Capabilities
Typical Inputs
Typical Outputs
What this skill does
- Design layered memory architectures
- Select memory frameworks
- Track entities across sessions
- Apply semantic retrieval
- Apply graph traversal
- Filter by temporal validity
- Consolidate accumulated memories
- Evaluate memory benchmarks
Inputs
- Agent persistence requirements
- Retrieval patterns
- Entity consistency needs
- Temporal validity requirements
- Memory benchmark data
- Infrastructure constraints
Outputs
- Memory architecture recommendations
- Framework selection guidance
- Retrieval strategy recommendations
- Consolidation plans
- Benchmark evaluation criteria