TensorStore Agent Memory (TSAM)
Google TensorStore-inspired Multi-Resolution Semantic Memory for Multi-Agent AI
Features
- 3D memory tensor (agent × time × embedding)
- Multi-resolution: recent, daily, weekly, all-time
- Semantic search (not grep)
- Agent similarity detection (SegCLR-inspired)
- Zero dependencies
Usage
from tensorstore_memory import AgentMemoryTensor
mem = AgentMemoryTensor()
mem.store("agent", "message", tags=["tag"])
results = mem.query("what happened?")