Vector Memory for OpenClaw

Zero-configuration smart memory search. Automatically uses neural embeddings when available, falls back to built-in search otherwise.

# Install and it just works
npx clawhub install vector-memory

# Optional: sync for better quality
node vector-memory/smart_memory.js --sync

✨ The Magic

Same function call. Automatic best method.

// This automatically does the right thing
memory_search("User principles values")

// If vector synced: finds "autonomy, competence, creation" (semantic!)
// If not synced: uses keyword search (fallback)

No configuration. No manual switching. No broken workflows.

🚀 Quick Start

From ClawHub (Recommended)

npx clawhub install vector-memory

Done. memory_search now has smart fallback.

From GitHub

curl -sL https://raw.githubusercontent.com/YOUR_USERNAME/vector-memory-openclaw/main/install.sh | bash

Manual

git clone https://github.com/YOUR_USERNAME/vector-memory-openclaw.git
cd vector-memory-openclaw/vector-memory && npm install

🎯 How It Works

User searches
      │
      ▼
┌─────────────┐
│ Vector ready?│
└──────┬──────┘
   Yes │    │ No
      ▼     ▼
┌────────┐ ┌──────────┐
│ Neural │ │ Keyword  │
│ Search │ │ Search   │
│ (best) │ │ (fast)   │
└────┬───┘ └────┬─────┘
     │          │
     └────┬─────┘
          ▼
    ┌────────────┐
    │  Results   │
    └────────────┘

Zero config philosophy:

1. Install → Works immediately (built-in fallback)
2. Sync → Gets better (vector embeddings)
3. Use → Always best available

📊 Before & After

🛠️ Usage

In OpenClaw

Just use memory_search:

const results = await memory_search("what we discussed", 5);
// Automatically uses best available method

CLI

# Search (auto-selects method)
node vector-memory/smart_memory.js --search "your query"

# Check what's active
node vector-memory/smart_memory.js --status

# Sync for better quality
node vector-memory/smart_memory.js --sync

📁 What's Included

vector-memory/
├── smart_memory.js           ← Main entry (auto-selects)
├── vector_memory_local.js    ← Neural embeddings
├── memory.js                 ← OpenClaw wrapper
├── package.json              ← Dependencies
└── references/
    ├── integration.md        ← Setup guide
    └── pgvector.md          ← Scale guide

skills/
└── vector-memory/
    ├── skill.json            ← OpenClaw manifest
    └── README.md             ← Skill docs

🔧 Requirements

• Node.js 18+
• ~80MB disk space (for model, cached after download)
• OpenClaw (or any Node.js agent)

🎛️ Tools

🔄 Auto-Sync (Optional)

Add to HEARTBEAT.md:

if [ -n "$(find memory MEMORY.md -newer vector-memory/.last_sync 2>/dev/null)" ]; then
    node vector-memory/smart_memory.js --sync && touch vector-memory/.last_sync
fi

📈 Performance

🐛 Troubleshooting

🧪 Verify Installation

node vector-memory/smart_memory.js --test

Checks: dependencies, vector index, search functionality, memory files.

📋 For Agent Developers

Add to your AGENTS.md:

## Memory Recall
Before answering about prior work, decisions, preferences:
1. Run memory_search with relevant query
2. Use memory_get for full context
3. If low confidence, say you checked

See full template in AGENTS.md.

🗂️ Suggested Memory Structure

workspace/
├── MEMORY.md              # Curated long-term memory
└── memory/
    ├── logs/              # Daily activity (YYYY-MM-DD.md)
    ├── projects/          # Project-specific notes
    ├── decisions/         # Important choices
    └── lessons/           # Mistakes learned

See MEMORY_STRUCTURE.md for templates.

🤝 Contributing

PRs welcome! See CONTRIBUTING.md (create one if needed).

📜 License

MIT

🙏 Acknowledgments

• Embeddings: Xenova Transformers
• Model: sentence-transformers/all-MiniLM-L6-v2
• Inspired by OpenClaw's memory system