Quickstart
From purchase to first memory stored in under 5 minutes. This guide covers everything from installing the package to running your first recall query.
Prerequisites
Node.js 18 or higher is required. Check your version:
node --version # must be v18.0.0 or higher
Step by Step
Purchase and get your licence key
Buy Vektor Slipstream at vektormemory.com/product. Your Polar licence key is emailed immediately after purchase. It looks like:
YOUR-LICENCE-KEY-HERE
Keys are UUID format — 32 hex characters in 8-4-4-4-12 groups. Copy it exactly as delivered.
Install the package
npm install vektor-slipstream
The package includes the ONNX embedding model (~23MB) — no separate download required. First install takes 30–60 seconds depending on your connection.
CLI Commands
Vektor Slipstream includes a full CLI via npx vektor:
npx vektor setup # Multi-app wizard: configures Claude Desktop, Cursor, Windsurf, VS Code, Continue, Groq Desktop npx vektor activate # Activate licence key (runs setup wizard automatically) npx vektor test # Test memory engine npx vektor status # System health check npx vektor mcp # Start MCP server (entry point for all apps) npx vektor rem # Run REM dream cycle npx vektor help # All commands
Windows users: if you see garbled characters, run [Console]::OutputEncoding = [System.Text.Encoding]::UTF8 in PowerShell first.
Set your licence key
First-run activation
When you call createMemory() for the first time, Slipstream activates your licence against Polar and caches the result to ~/.vektor/licence.json. You will not be prompted again on that machine.
There are three ways to provide your key:
# .env file or shell export
VEKTOR_LICENCE_KEY=YOUR-LICENCE-KEY-HERE
const memory = await createMemory({
agentId: 'my-agent',
licenceKey: process.env.VEKTOR_LICENCE_KEY,
});
╔══════════════════════════════════════════════════════╗
║ VEKTOR SLIPSTREAM — LICENCE REQUIRED ║
╚══════════════════════════════════════════════════════╝
Paste your licence key from your Polar purchase email:
Paste your licence key › YOUR-LICENCE-KEY-HERE
✓ Licence validated — this machine is now activated.
Your key has been saved to ~/.vektor/licence.json
Machine limit — 3 activations per licence
Each licence activates on up to 3 machines. If you hit the limit, deactivate an old machine first:
node -e "require('vektor-slipstream/vektor-licence').deactivateMachine(process.env.VEKTOR_LICENCE_KEY)"
Or manage activations directly at polar.sh → your customer portal. Enterprise licences (10 seats) available — contact hello@vektormemory.com.
Add your key to your environment. Create a .env file in your project root:
VEKTOR_LICENCE_KEY=YOUR-LICENCE-KEY-HERE
Or pass it directly in code (not recommended for production):
const memory = await createMemory({ agentId: 'my-agent', licenceKey: 'YOUR-LICENCE-KEY-HERE', // or use env var });
Store your first memory
Create test-memory.js and run it:
const { createMemory } = require('vektor-slipstream'); async function main() { const memory = await createMemory({ agentId: 'my-agent', licenceKey: process.env.VEKTOR_LICENCE_KEY, dbPath: './memory.db', }); // Store a memory const { id } = await memory.remember('User prefers TypeScript over JavaScript'); console.log('Stored memory id:', id); // Recall by semantic similarity const results = await memory.recall('coding preferences', 5); console.log('Recalled:', results); } main().catch(console.error);
node test-memory.js
Expected output
╔══════════════════════════════════════════════════════╗ ║ VEKTOR SLIPSTREAM — ACTIVE ║ ╚══════════════════════════════════════════════════════╝ ⚙️ EP: CPU (WASM SIMD) 🧠 Model: all-MiniLM-L6-v2 INT8 quantized ⚡ Embed: 24ms (post-warmup) 💾 DB: WAL | mmap:1GB | cache:64MB 🔥 Warm: ✓ ⏱ Boot: 312ms total Stored memory id: 1 Recalled: [ { id: 1, content: 'User prefers TypeScript over JavaScript', score: 0.97, importance: 1 } ]
Next Steps
You now have a working memory layer. Here's where to go next:
- → API Reference — full documentation for every method
- → LangChain integration — add persistent memory to your LangChain agent
- → OpenAI Agents SDK — wrap your OpenAI agent with memory
- → Claude MCP — connect Claude Desktop to Vektor memory