neuralmemory 1.6.0 → 1.6.2

package/README.md

@@ -0,0 +1,139 @@
+# NeuralMemory — OpenClaw Plugin
+
+Brain-inspired persistent memory for AI agents. Stores experiences as interconnected neurons and recalls them through spreading activation, mimicking how the human brain works.
+
+This is the **OpenClaw plugin** for [Neural Memory](https://github.com/nhadaututheky/neural-memory).
+
+## Prerequisites
+
+```bash
+pip install neural-memory
+```
+
+Python 3.11+ required. Verify the install:
+
+```bash
+nmem-mcp --help
+```
+
+## Install
+
+```bash
+npm install neuralmemory
+```
+
+Or add to your OpenClaw config directly.
+
+## OpenClaw Setup
+
+Add to `~/.openclaw/openclaw.json`:
+
+```json
+{
+  "plugins": {
+    "slots": {
+      "memory": "neuralmemory"
+    },
+    "entries": {
+      "neuralmemory": {
+        "config": {
+          "pythonPath": "python",
+          "brain": "default",
+          "autoContext": true,
+          "autoCapture": true
+        }
+      }
+    }
+  }
+}
+```
+
+> **Important**: Setting `slots.memory = "neuralmemory"` disables the default `memory-core` plugin. Without this, agents may still use `memory_search` instead of NeuralMemory tools.
+
+## Tools
+
+The plugin exposes 6 core tools to OpenClaw agents:
+
+| Tool | Description |
+|------|-------------|
+| `nmem_remember` | Store a memory (auto-detects type: fact, decision, insight, etc.) |
+| `nmem_recall` | Search memories via spreading activation |
+| `nmem_context` | Get recent context for the current session |
+| `nmem_todo` | Quick TODO shortcut |
+| `nmem_stats` | Brain statistics (memory count, freshness) |
+| `nmem_health` | Brain health diagnostics (purity score, warnings) |
+
+## Configuration
+
+| Option | Type | Default | Description |
+|--------|------|---------|-------------|
+| `pythonPath` | string | `"python"` | Python executable with `neural-memory` installed |
+| `brain` | string | `"default"` | Brain name (each workspace can have its own) |
+| `autoContext` | boolean | `true` | Auto-inject relevant memories before each agent run |
+| `autoCapture` | boolean | `true` | Auto-extract memories after each agent run |
+| `contextDepth` | integer | `1` | Recall depth: 0=instant, 1=context, 2=habits, 3=deep |
+| `maxContextTokens` | integer | `500` | Max tokens for auto-context injection |
+| `timeout` | integer | `30000` | MCP request timeout (ms) |
+
+## How It Works
+
+```
+OpenClaw Agent
+    |
+    v
+NeuralMemory Plugin (this package)
+    |  Spawns + manages lifecycle
+    v
+nmem-mcp (Python MCP server, stdio transport)
+    |
+    v
+~/.neuralmemory/brains/<brain>.db (SQLite)
+```
+
+The plugin spawns `nmem-mcp` as a subprocess and communicates via JSON-RPC over stdio. Memories are stored in a local SQLite database.
+
+## Troubleshooting
+
+**Timeout on startup**: If you see `MCP timeout: initialize (30000ms)`, the Python process is slow to start. Fix:
+
+```bash
+# Pre-install to avoid cold start delays
+pip install neural-memory
+
+# Or increase the timeout in your config
+"timeout": 60000
+```
+
+**"nmem-mcp not found"**: Ensure `neural-memory` is installed in the Python environment that `pythonPath` points to.
+
+**Schema validation errors**: Upgrade to `neural-memory>=2.28.0` which fixes empty schema validation for strict JSON Schema validators.
+
+## Full Feature Set
+
+The underlying Neural Memory system provides 39 MCP tools including:
+
+- Spreading activation recall (not keyword search)
+- Memory maturation (STM -> Working -> Episodic -> Semantic)
+- Trust scores with source-based ceiling caps
+- Batch remember (up to 20 memories per call)
+- Cognitive reasoning (hypotheses, evidence, predictions)
+- Knowledge base training from PDF/DOCX/PPTX/HTML/JSON/CSV
+- Multi-device sync with conflict resolution
+- Fernet encryption for sensitive content
+- React dashboard for visual brain exploration
+
+Access the full toolset via [Claude Code](https://claude.com/claude-code) or any MCP-compatible client:
+
+```bash
+claude mcp add --scope user neural-memory -- nmem-mcp
+```
+
+## Links
+
+- [Neural Memory on GitHub](https://github.com/nhadaututheky/neural-memory)
+- [Neural Memory on PyPI](https://pypi.org/project/neural-memory/)
+- [Documentation](https://nhadaututheky.github.io/neural-memory/)
+
+## License
+
+MIT

package/openclaw.plugin.json

@@ -3,7 +3,7 @@
   "kind": "memory",
   "name": "NeuralMemory",
   "description": "Brain-inspired persistent memory for AI agents — neurons, synapses, and fibers. REQUIRED: set plugins.slots.memory = \"neuralmemory\" in openclaw.json to disable the default memory-core plugin and activate NeuralMemory as the exclusive memory provider.",
-  "version": "1.5.0",
+  "version": "1.6.0",
   "configSchema": {
     "jsonSchema": {
       "type": "object",

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "neuralmemory",
-  "version": "1.6.0",
+  "version": "1.6.2",
   "description": "NeuralMemory plugin for OpenClaw — brain-inspired persistent memory for AI agents",
   "type": "module",
   "main": "dist/index.js",
@@ -30,10 +30,12 @@
     "ai-agent",
     "persistent-memory"
   ],
+  "author": "NeuralMemory Contributors",
   "license": "MIT",
+  "homepage": "https://github.com/nhadaututtheky/neural-memory#readme",
   "repository": {
     "type": "git",
-    "url": "https://github.com/pinkponk/NeuralMemory"
+    "url": "https://github.com/nhadaututtheky/neural-memory"
   },
   "files": [
     "src/",