Neuro-Memory Studio v0.1.0

Enter a partial cue or query concept. The CA3 recurrent collateral network reconstructs the full memory trace even from fragmented inputs, while CA1 performs sparse separation.

During offline consolidation (sleep cycles), high-frequency associative pathways are stabilized into long-term cortical memory, while low-importance, degraded synaptic traces are pruned according to Hermann Ebbinghaus's exponential forgetting law: R = exp(-t / S).

Connect Claude Desktop, Cursor, Cline, OpenClaw, or custom AI agent swarms directly to the persistent synaptic memory substrate.

{
  "mcpServers": {
    "neuro-memory": {
      "command": "neuro-memory-daemon",
      "args": ["mcp"]
    }
  }
}

{
  "name": "neuro-memory",
  "command": "python3",
  "args": ["-m", "neuro_memory_daemon.cli", "mcp"]
}

# Store Memory
curl -X POST http://127.0.0.1:8788/api/memories \
  -H "Content-Type: application/json" \
  -d '{"text": "Refactored UI to Material 3", "tags": ["ui", "refactor"], "category": "architecture"}'

# Associative Recall
curl -X POST http://127.0.0.1:8788/api/recall \
  -H "Content-Type: application/json" \
  -d '{"query": "Material 3 design system", "top_k": 5}'